Skip to content

Plugins

Plugins#

Source stage#

FROM-ADAMS-IC#

Reads image classification annotations in the ADAMS report-format

Domain(s):#
  • Image Classification Domain
Options:#
usage: from-adams-ic [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                     [--seed SEED] [-e FORMAT FORMAT FORMAT] -c FIELD

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  -e FORMAT FORMAT FORMAT, --extensions FORMAT FORMAT FORMAT
                        image format extensions in order of preference (default: [<ImageFormat.PNG:
                        (frozenset({'png', 'PNG'}), 'PNG')>, <ImageFormat.JPG: (frozenset({'jpg',
                        'JPEG', 'jpeg', 'JPG'}), 'JPEG')>, <ImageFormat.BMP: (frozenset({'BMP',
                        'bmp'}), 'BMP')>])
  -c FIELD, --class-field FIELD
                        the report field containing the image class (default: None)

FROM-ADAMS-OD#

Reads image object-detection annotations in the ADAMS report-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-adams-od [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                     [--seed SEED] [-e FORMAT FORMAT FORMAT] [-p PREFIXES [PREFIXES ...]]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  -e FORMAT FORMAT FORMAT, --extensions FORMAT FORMAT FORMAT
                        image format extensions in order of preference (default: [<ImageFormat.PNG:
                        (frozenset({'png', 'PNG'}), 'PNG')>, <ImageFormat.JPG: (frozenset({'jpg',
                        'JPEG', 'jpeg', 'JPG'}), 'JPEG')>, <ImageFormat.BMP: (frozenset({'BMP',
                        'bmp'}), 'BMP')>])
  -p PREFIXES [PREFIXES ...], --prefixes PREFIXES [PREFIXES ...]
                        prefixes to parse (default: [])

FROM-AUDIO-FILES-AC#

Dummy reader that turns audio files into a classification dataset.

Domain(s):#
  • Audio classification domain
Options:#
usage: from-audio-files-ac [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                           [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-AUDIO-FILES-SP#

Dummy reader that turns audio files into a speech dataset.

Domain(s):#
  • Speech Domain
Options:#
usage: from-audio-files-sp [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                           [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-BLUE-CHANNEL-IS#

Reads image segmentation files in the blue-channel format

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: from-blue-channel-is [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                            [--seed SEED] [--image-path-rel PATH] --labels LABEL [LABEL ...]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  --image-path-rel PATH
                        Relative path to image files from annotations (default: .)
  --labels LABEL [LABEL ...]
                        specifies the labels for each index (default: None)

FROM-COCO-OD#

Reads image object-detection annotations in the MS-COCO JSON-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-coco-od [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                    [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-COMMON-VOICE-SP#

Reads speech transcriptions in the Mozilla Common-Voice TSV-format

Domain(s):#
  • Speech Domain
Options:#
usage: from-common-voice-sp [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                            [--seed SEED] [--rel-path REL_PATH]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  --rel-path REL_PATH   the relative path from the annotations file to the audio files (default: .)

FROM-COQUI-STT-SP#

Reads speech transcriptions in the Coqui STT CSV-format

Domain(s):#
  • Speech Domain
Options:#
usage: from-coqui-stt-sp [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                         [--seed SEED] [--rel-path REL_PATH]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  --rel-path REL_PATH   the relative path from the annotations file to the audio files (default: .)

FROM-COQUI-TTS-SP#

Reads speech transcriptions in the Coqui TTS text-format

Domain(s):#
  • Speech Domain
Options:#
usage: from-coqui-tts-sp [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                         [--seed SEED] [--rel-path REL_PATH]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  --rel-path REL_PATH   the relative path from the annotations file to the audio files (default: .)

FROM-FESTVOX-SP#

Reads speech transcriptions in the Festival FestVox format

Domain(s):#
  • Speech Domain
Options:#
usage: from-festvox-sp [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                       [--seed SEED] [--rel-path REL_PATH]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  --rel-path REL_PATH   the relative path from the annotations file to the audio files (default: .)

FROM-GRAYSCALE-IS#

Reads image segmentation files in the grayscale format

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: from-grayscale-is [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                         [--seed SEED] [--image-path-rel PATH] --labels LABEL [LABEL ...]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  --image-path-rel PATH
                        Relative path to image files from annotations (default: .)
  --labels LABEL [LABEL ...]
                        specifies the labels for each index (default: None)

FROM-IMAGES-IC#

Dummy reader that turns images into an image classification dataset.

Domain(s):#
  • Image Classification Domain
Options:#
usage: from-images-ic [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                      [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-IMAGES-IS#

Dummy reader that turns images into an image segmentation dataset.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: from-images-is [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                      [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-IMAGES-OD#

Dummy reader that turns images into an object detection dataset.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-images-od [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                      [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-INDEXED-PNG-IS#

Reads image segmentation files in the indexed-PNG format

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: from-indexed-png-is [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                           [--seed SEED] [--image-path-rel PATH] --labels LABEL [LABEL ...]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  --image-path-rel PATH
                        Relative path to image files from annotations (default: .)
  --labels LABEL [LABEL ...]
                        specifies the labels for each index (default: None)

FROM-LAYER-SEGMENTS-IS#

Reads in the layer-segments image-segmentation format from disk, where each label has a binary PNG storing the mask for that label

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: from-layer-segments-is [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                              [--seed SEED] [--invert] [--label-separator SEPARATOR] --labels LABEL
                              [LABEL ...] [--lenient] [--image-path-rel PATH]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  --invert              inverts the colors in the annotations (b/w <-> w/b) (default: False)
  --label-separator SEPARATOR
                        the separator between the base filename and the label (default: -)
  --labels LABEL [LABEL ...]
                        specifies the labels for each index (default: None)
  --lenient             converts non-binary images with only two unique colors into binary ones
                        rather than throwing an exception (default: False)
  --image-path-rel PATH
                        Relative path to image files from annotations (default: .)

FROM-OPEX-OD#

Reads image object-detection annotations in the OPEX format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-opex-od [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                    [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-ROI-OD#

Reads image object-detection annotations in the ROI CSV-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-roi-od [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                   [--seed SEED] [-e FORMAT FORMAT FORMAT] [--prefix READER_PREFIX]
                   [--suffix READER_SUFFIX]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  -e FORMAT FORMAT FORMAT, --extensions FORMAT FORMAT FORMAT
                        image format extensions in order of preference (default: [<ImageFormat.PNG:
                        (frozenset({'png', 'PNG'}), 'PNG')>, <ImageFormat.JPG: (frozenset({'jpg',
                        'JPEG', 'jpeg', 'JPG'}), 'JPEG')>, <ImageFormat.BMP: (frozenset({'BMP',
                        'bmp'}), 'BMP')>])
  --prefix READER_PREFIX
                        the prefix for output filenames (default = '') (default: None)
  --suffix READER_SUFFIX
                        the suffix for output filenames (default = '-rois.csv') (default: None)

FROM-SUBDIR-AC#

Reads audio files from sub-directories named after their class labels.

Domain(s):#
  • Audio classification domain
Options:#
usage: from-subdir-ac [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                      [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-SUBDIR-IC#

Reads images from sub-directories named after their class labels.

Domain(s):#
  • Image Classification Domain
Options:#
usage: from-subdir-ic [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                      [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-TF-OD#

Reads image object-detection annotations in the TFRecords binary format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-tf-od [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                  [--seed SEED] [--mask-threshold THRESHOLD] [--sample-stride STRIDE]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  --mask-threshold THRESHOLD
                        the threshold to use when calculating polygons from masks (default: 0.9)
  --sample-stride STRIDE
                        the stride to use when calculating polygons from masks (default: 1)

FROM-VGG-OD#

Reads image object-detection annotations in the VGG JSON-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-vgg-od [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                   [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-VIDEO-FILE-OD#

Reads frames from a video file.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-video-file-od [-f FROM_FRAME] [-i INPUT_FILE] [-m MAX_FRAMES] [-n NTH_FRAME] [-p PREFIX]
                          [-t TO_FRAME]

optional arguments:
  -f FROM_FRAME, --from-frame FROM_FRAME
                        determines with which frame to start the stream (1-based index) (default: 1)
  -i INPUT_FILE, --input INPUT_FILE
                        the video file to read (default: )
  -m MAX_FRAMES, --max-frames MAX_FRAMES
                        determines the maximum number of frames to read; ignored if <=0 (default:
                        -1)
  -n NTH_FRAME, --nth-frame NTH_FRAME
                        determines whether frames get skipped and only evert nth frame gets
                        forwarded (default: 1)
  -p PREFIX, --prefix PREFIX
                        the prefix to use for the frames (default: )
  -t TO_FRAME, --to-frame TO_FRAME
                        determines after which frame to stop (1-based index); ignored if <=0
                        (default: -1)

FROM-VOC-OD#

Reads image object-detection annotations in the Pascal VOC XML-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-voc-od [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                   [--seed SEED]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)

FROM-WEBCAM-OD#

Reads frames from a webcam.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-webcam-od [-f FROM_FRAME] [-m MAX_FRAMES] [-n NTH_FRAME] [-p PREFIX] [-t TO_FRAME]
                      [-i WEBCAM_ID]

optional arguments:
  -f FROM_FRAME, --from-frame FROM_FRAME
                        determines with which frame to start the stream (1-based index) (default: 1)
  -m MAX_FRAMES, --max-frames MAX_FRAMES
                        determines the maximum number of frames to read; ignored if <=0 (default:
                        -1)
  -n NTH_FRAME, --nth-frame NTH_FRAME
                        determines whether frames get skipped and only evert nth frame gets
                        forwarded (default: 1)
  -p PREFIX, --prefix PREFIX
                        the prefix to use for the frames (default: webcam-)
  -t TO_FRAME, --to-frame TO_FRAME
                        determines after which frame to stop (1-based index); ignored if <=0
                        (default: -1)
  -i WEBCAM_ID, --webcam-id WEBCAM_ID
                        the webcam ID to read from (default: 0)

FROM-YOLO-OD#

Reads image object-detection annotations in the YOLO format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: from-yolo-od [-I FILENAME] [-i FILENAME] [-N FILENAME] [-n FILENAME] [-o FILENAME]
                    [--seed SEED] [--image-path-rel PATH] [-l PATH]

optional arguments:
  -I FILENAME, --inputs-file FILENAME
                        Files containing lists of input files (can use glob syntax) (default: [])
  -i FILENAME, --input FILENAME
                        Input files (can use glob syntax) (default: [])
  -N FILENAME, --negatives-file FILENAME
                        Files containing lists of negative files (can use glob syntax) (default: [])
  -n FILENAME, --negative FILENAME
                        Files that have no annotations (can use glob syntax) (default: [])
  -o FILENAME, --output-file FILENAME
                        optional file to write read filenames into (default: None)
  --seed SEED           the seed to use for randomisation (default: None)
  --image-path-rel PATH
                        Relative path to image files from annotations (default: None)
  -l PATH, --labels PATH
                        Path to the labels file (default: None)

GENERIC-SOURCE-AC#

Generic audio classification source.

Domain(s):#
  • Audio classification domain
Options:#
usage: generic-source-ac [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-SOURCE-IC#

Generic image classification source.

Domain(s):#
  • Image Classification Domain
Options:#
usage: generic-source-ic [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-SOURCE-IS#

Generic image segmentation source.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: generic-source-is [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-SOURCE-OD#

Generic object detection source.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: generic-source-od [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-SOURCE-SP#

Generic speech source.

Domain(s):#
  • Speech Domain
Options:#
usage: generic-source-sp [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

Processor stage#

ADD-ANNOTATION-OVERLAY-IC#

Adds the image classification label on top of images passing through.

Domain(s):#
  • Image Classification Domain
Options:#
usage: add-annotation-overlay-ic [--background-color BACKGROUND_COLOR]
                                 [--background-margin BACKGROUND_MARGIN] [--fill-background]
                                 [--font-color FONT_COLOR] [--font-family FONT_FAMILY]
                                 [--font-size FONT_SIZE] [--position TEXT_PLACEMENT]

optional arguments:
  --background-color BACKGROUND_COLOR
                        the RGB color triplet to use for the background. (default: 0,0,0)
  --background-margin BACKGROUND_MARGIN
                        the margin in pixels around the background. (default: 2)
  --fill-background     whether to fill the background of the text with the specified color.
                        (default: False)
  --font-color FONT_COLOR
                        the RGB color triplet to use for the font. (default: 255,255,255)
  --font-family FONT_FAMILY
                        the name of the TTF font-family to use, note: any hyphens need escaping with
                        backslash. (default: sans\-serif)
  --font-size FONT_SIZE
                        the size of the font. (default: 14)
  --position TEXT_PLACEMENT
                        the position of the label (X,Y). (default: 5,5)

ADD-ANNOTATION-OVERLAY-IS#

Adds the image segmentation annotations on top of images passing through.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: add-annotation-overlay-is [--alpha ALPHA] [--colors COLORS [COLORS ...]]
                                 [--labels LABELS [LABELS ...]]

optional arguments:
  --alpha ALPHA         the alpha value to use for overlaying the annotations (0: transparent, 255:
                        opaque). (default: 64)
  --colors COLORS [COLORS ...]
                        the RGB triplets (R,G,B) of custom colors to use, uses default colors if not
                        supplied (default: [])
  --labels LABELS [LABELS ...]
                        the labels of annotations to overlay, overlays all if omitted (default: [])

ADD-ANNOTATION-OVERLAY-OD#

Adds object detection overlays to images passing through.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: add-annotation-overlay-od [--colors COLORS [COLORS ...]] [--fill] [--fill-alpha FILL_ALPHA]
                                 [--font-family FONT_FAMILY] [--font-size FONT_SIZE] [--force-bbox]
                                 [--label-key LABEL_KEY] [--labels LABELS [LABELS ...]]
                                 [--num-decimals NUM_DECIMALS] [--outline-alpha OUTLINE_ALPHA]
                                 [--outline-thickness OUTLINE_THICKNESS] [--text-format TEXT_FORMAT]
                                 [--text-placement TEXT_PLACEMENT] [--vary-colors]

optional arguments:
  --colors COLORS [COLORS ...]
                        the RGB triplets (R,G,B) of custom colors to use, uses default colors if not
                        supplied (default: [])
  --fill                whether to fill the bounding boxes/polygons (default: False)
  --fill-alpha FILL_ALPHA
                        the alpha value to use for the filling (0: transparent, 255: opaque).
                        (default: 128)
  --font-family FONT_FAMILY
                        the name of the TTF font-family to use, note: any hyphens need escaping with
                        backslash. (default: sans\-serif)
  --font-size FONT_SIZE
                        the size of the font. (default: 14)
  --force-bbox          whether to force a bounding box even if there is a polygon available
                        (default: False)
  --label-key LABEL_KEY
                        the key in the meta-data that contains the label. (default: type)
  --labels LABELS [LABELS ...]
                        the labels of annotations to overlay, overlays all if omitted (default: [])
  --num-decimals NUM_DECIMALS
                        the number of decimals to use for float numbers in the text format string.
                        (default: 3)
  --outline-alpha OUTLINE_ALPHA
                        the alpha value to use for the outline (0: transparent, 255: opaque).
                        (default: 255)
  --outline-thickness OUTLINE_THICKNESS
                        the line thickness to use for the outline, <1 to turn off. (default: 3)
  --text-format TEXT_FORMAT
                        template for the text to print on top of the bounding box or polygon, '{PH}'
                        is a placeholder for the 'PH' value from the meta-data or 'label' for the
                        current label; ignored if empty. (default: {label})
  --text-placement TEXT_PLACEMENT
                        comma-separated list of vertical (T=top, C=center, B=bottom) and horizontal
                        (L=left, C=center, R=right) anchoring. (default: T,L)
  --vary-colors         whether to vary the colors of the outline/filling regardless of label
                        (default: False)

CHECK-DUPLICATE-FILENAMES#

Causes the conversion stream to halt when multiple dataset items have the same filename

Domain(s):#
  • Speech Domain
  • Image Object-Detection Domain
  • Image Classification Domain
  • Image Segmentation Domain
  • Audio classification domain
Options:#
usage: check-duplicate-filenames

CLEAN-TRANSCRIPT#

ISP that cleans speech transcripts.

Domain(s):#
  • Speech Domain
Options:#
usage: clean-transcript [-b] [-c CUSTOM] [-a] [-l] [-n] [-p] [-q] [--verbose]

optional arguments:
  -b, --brackets        removes brackets: ()[]{}〈〉 (default: False)
  -c CUSTOM, --custom CUSTOM
                        the custom characters to remove (default: )
  -a, --non-alpha-numeric
                        removes all characters that are not alpha-numeric (default: False)
  -l, --non-letters     removes all characters that are not letters (default: False)
  -n, --numeric         removes all numeric characters (default: False)
  -p, --punctuation     removes punctuation characters: :;,.!? (default: False)
  -q, --quotes          removes quotes: '"‘’“”‹›«» (default: False)
  --verbose             outputs information about processed transcripts (default: False)

COERCE-BOX#

Converts all annotation bounds into box regions

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: coerce-box

COERCE-MASK#

Converts all annotation bounds into polygon regions

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: coerce-mask

COMBINE-ANNOTATIONS-OD#

Combines object detection annotations from images passing through into a single annotation.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: combine-annotations-od [--combination COMBINATION] [--min-iou MIN_IOU]

optional arguments:
  --combination COMBINATION
                        how to combine the annotations (union|intersect); the 'stream_index' key in
                        the meta-data contains the stream index (default: intersect)
  --min-iou MIN_IOU     the minimum IoU (intersect over union) to use for identifying objects that
                        overlap (default: 0.7)

CONVERT-IMAGE-FORMAT#

Converts images from one format to another

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
  • Image Segmentation Domain
Options:#
usage: convert-image-format -f FORMAT

optional arguments:
  -f FORMAT, --format FORMAT
                        format to convert images to (default: None)

CONVERT-TO-MONO#

Converts audio files to monophonic.

Domain(s):#
  • Speech Domain
  • Audio classification domain
Options:#
usage: convert-to-mono

CONVERT-TO-WAV#

Converts mp3/flac/ogg to wav.

Domain(s):#
  • Speech Domain
  • Audio classification domain
Options:#
usage: convert-to-wav [-s SAMPLE_RATE]

optional arguments:
  -s SAMPLE_RATE, --sample-rate SAMPLE_RATE
                        the sample rate to use for the audio data, for overriding the native rate.
                        (default: None)

CROP#

Crops images.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
Options:#
usage: crop [-m IMGAUG_MODE] [--suffix IMGAUG_SUFFIX] [-f PERCENT_FROM] [-t PERCENT_TO] [-s SEED]
            [-a] [-T THRESHOLD] [-u]

optional arguments:
  -m IMGAUG_MODE, --mode IMGAUG_MODE
                        the image augmentation mode to use, available modes: replace, add (default:
                        replace)
  --suffix IMGAUG_SUFFIX
                        the suffix to use for the file names in case of augmentation mode add
                        (default: None)
  -f PERCENT_FROM, --from-percent PERCENT_FROM
                        the minimum percent to crop from images (default: None)
  -t PERCENT_TO, --to-percent PERCENT_TO
                        the maximum percent to crop from images (default: None)
  -s SEED, --seed SEED  the seed value to use for the random number generator; randomly seeded if
                        not provided (default: None)
  -a, --seed-augmentation
                        whether to seed the augmentation; if specified, uses the seeded random
                        generator to produce a seed value from 0 to 1000 for the augmentation.
                        (default: False)
  -T THRESHOLD, --threshold THRESHOLD
                        the threshold to use for Random.rand(): if equal or above, augmentation gets
                        applied; range: 0-1; default: 0 (= always) (default: None)
  -u, --update-size     whether to update the image size after the crop operation or scale back to
                        original size (default: False)

DIMENSION-DISCARDER#

Removes annotations which fall outside certain size constraints

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: dimension-discarder [--max-area MAX_AREA] [--max-height MAX_HEIGHT] [--max-width MAX_WIDTH]
                           [--min-area MIN_AREA] [--min-height MIN_HEIGHT] [--min-width MIN_WIDTH]
                           [--verbose]

optional arguments:
  --max-area MAX_AREA   the maximum area of annotations to convert (default: None)
  --max-height MAX_HEIGHT
                        the maximum height of annotations to convert (default: None)
  --max-width MAX_WIDTH
                        the maximum width of annotations to convert (default: None)
  --min-area MIN_AREA   the minimum area of annotations to convert (default: None)
  --min-height MIN_HEIGHT
                        the minimum height of annotations to convert (default: None)
  --min-width MIN_WIDTH
                        the minimum width of annotations to convert (default: None)
  --verbose             outputs information when discarding annotations (default: False)

DISCARD-INVALID-IMAGES#

Discards images that cannot be loaded (e.g., corrupt image file or annotations with no image)

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
  • Image Segmentation Domain
Options:#
usage: discard-invalid-images [-v]

optional arguments:
  -v, --verbose  whether to output debugging information (default: False)

DISCARD-NEGATIVES#

Discards negative examples (those without annotations) from the stream

Domain(s):#
  • Speech Domain
  • Image Object-Detection Domain
  • Image Classification Domain
  • Image Segmentation Domain
  • Audio classification domain
Options:#
usage: discard-negatives

DROP-FRAMES#

Drops frames from the stream.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
  • Image Segmentation Domain
Options:#
usage: drop-frames [-n NTH_FRAME]

optional arguments:
  -n NTH_FRAME, --nth-frame NTH_FRAME
                        which nth frame to drop, e..g, '2' means to drop every 2nd frame; passes
                        frames through if <=1 (default: 0)

FILTER-FRAMES-BY-LABEL-OD#

Filters frames from the stream using the labels in the annotations, i.e., keeps or drops frames depending on presence/absence of labels.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
  • Image Segmentation Domain
Options:#
usage: filter-frames-by-label-od [--excluded-labels EXCLUDED_LABELS] [--key-label KEY_LABEL]
                                 [--key-score KEY_SCORE] [--min-score MIN_SCORE]
                                 [--required-labels REQUIRED_LABELS] [-v]

optional arguments:
  --excluded-labels EXCLUDED_LABELS
                        the comma-separated list of labels that will automatically drop the frame
                        when present in the frame (default: )
  --key-label KEY_LABEL
                        the meta-data key in the annotations that contains the label. (default:
                        type)
  --key-score KEY_SCORE
                        the meta-data key in the annotations to use for storing the prediction
                        score. (default: score)
  --min-score MIN_SCORE
                        the minimum score that predictions must have in order to be included in the
                        label checks, ignored if not supplied (default: None)
  --required-labels REQUIRED_LABELS
                        the comma-separated list of labels that must be present in the frame,
                        otherwise it gets dropped (default: )
  -v, --verbose         whether to output debugging information. (default: False)

FILTER-LABELS#

Filters detected objects down to those with specified labels or, in case of image classification, removes the label if it doesn't match.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
Options:#
usage: filter-labels [-l LABELS [LABELS ...]] [--min-iou FLOAT] [-r regexp] [--region x,y,w,h]

optional arguments:
  -l LABELS [LABELS ...], --labels LABELS [LABELS ...]
                        labels to use (default: [])
  --min-iou FLOAT       the minimum IoU (intersect over union) that the object must have with the
                        region in order to be considered an overlap (object detection only)
                        (default: 0.01)
  -r regexp, --regexp regexp
                        regular expression for using only a subset of labels (default: None)
  --region x,y,w,h      region that the object must overlap with in order to be included (object
                        detection only). Between 0-1 the values are considered normalized, otherwise
                        absolute pixels. (default: None)

FILTER-METADATA#

Filters detected objects based on their meta-data.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: filter-metadata [-c COMPARISON] [-k KEY] [-t VALUE_TYPE]

optional arguments:
  -c COMPARISON, --comparison COMPARISON
                        the comparison to apply to the value: for bool/numeric/string '=OTHER' and
                        '!=OTHER' can be used, for numeric furthermore '<OTHER', '<=OTHER',
                        '>=OTHER', '>OTHER'. E.g.: '<3.0' for numeric types will discard any
                        annotations that have a value of 3.0 or larger (default: None)
  -k KEY, --key KEY     the key of the meta-data value to use for the filtering (default: None)
  -t VALUE_TYPE, --value-type VALUE_TYPE
                        the data type that the value represents, available options:
                        bool|numeric|string (default: None)

FLIP#

Flips images either left-to-right, up-to-down or both.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
Options:#
usage: flip [-d DIRECTION] [-m IMGAUG_MODE] [--suffix IMGAUG_SUFFIX] [-s SEED] [-a] [-T THRESHOLD]

optional arguments:
  -d DIRECTION, --direction DIRECTION
                        the direction to flip, available options: lr, up, lrup (default: None)
  -m IMGAUG_MODE, --mode IMGAUG_MODE
                        the image augmentation mode to use, available modes: replace, add (default:
                        replace)
  --suffix IMGAUG_SUFFIX
                        the suffix to use for the file names in case of augmentation mode add
                        (default: None)
  -s SEED, --seed SEED  the seed value to use for the random number generator; randomly seeded if
                        not provided (default: None)
  -a, --seed-augmentation
                        whether to seed the augmentation; if specified, uses the seeded random
                        generator to produce a seed value from 0 to 1000 for the augmentation.
                        (default: False)
  -T THRESHOLD, --threshold THRESHOLD
                        the threshold to use for Random.rand(): if equal or above, augmentation gets
                        applied; range: 0-1; default: 0 (= always) (default: None)

GAUSSIAN-BLUR#

Applies gaussian blur to images.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
Options:#
usage: gaussian-blur [-m IMGAUG_MODE] [--suffix IMGAUG_SUFFIX] [-s SEED] [-a] [-f SIGMA_FROM]
                     [-t SIGMA_TO] [-T THRESHOLD]

optional arguments:
  -m IMGAUG_MODE, --mode IMGAUG_MODE
                        the image augmentation mode to use, available modes: replace, add (default:
                        replace)
  --suffix IMGAUG_SUFFIX
                        the suffix to use for the file names in case of augmentation mode add
                        (default: None)
  -s SEED, --seed SEED  the seed value to use for the random number generator; randomly seeded if
                        not provided (default: None)
  -a, --seed-augmentation
                        whether to seed the augmentation; if specified, uses the seeded random
                        generator to produce a seed value from 0 to 1000 for the augmentation.
                        (default: False)
  -f SIGMA_FROM, --from-sigma SIGMA_FROM
                        the minimum sigma for the blur to apply to the images (default: None)
  -t SIGMA_TO, --to-sigma SIGMA_TO
                        the maximum sigma for the blur to apply to the images (default: None)
  -T THRESHOLD, --threshold THRESHOLD
                        the threshold to use for Random.rand(): if equal or above, augmentation gets
                        applied; range: 0-1; default: 0 (= always) (default: None)

GENERIC-ISP-AC#

Generic audio classification ISP.

Domain(s):#
  • Audio classification domain
Options:#
usage: generic-isp-ac [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-ISP-IC#

Generic image classification ISP.

Domain(s):#
  • Image Classification Domain
Options:#
usage: generic-isp-ic [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-ISP-IS#

Generic image segmentation ISP.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: generic-isp-is [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-ISP-OD#

Generic object detection ISP.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: generic-isp-od [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-ISP-SP#

Generic speech ISP.

Domain(s):#
  • Speech Domain
Options:#
usage: generic-isp-sp [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

HSL-GRAYSCALE#

Turns RGB images into fake grayscale ones by converting them to HSL and then using the L channel for all channels. The brightness can be influenced and varied even.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
Options:#
usage: hsl-grayscale [-f FACTOR_FROM] [-t FACTOR_TO] [-m IMGAUG_MODE] [--suffix IMGAUG_SUFFIX]
                     [-s SEED] [-a] [-T THRESHOLD]

optional arguments:
  -f FACTOR_FROM, --from-factor FACTOR_FROM
                        the start of the factor range to apply to the L channel to darken or lighten
                        the image (<1: darker, >1: lighter) (default: None)
  -t FACTOR_TO, --to-factor FACTOR_TO
                        the end of the factor range to apply to the L channel to darken or lighten
                        the image (<1: darker, >1: lighter) (default: None)
  -m IMGAUG_MODE, --mode IMGAUG_MODE
                        the image augmentation mode to use, available modes: replace, add (default:
                        replace)
  --suffix IMGAUG_SUFFIX
                        the suffix to use for the file names in case of augmentation mode add
                        (default: None)
  -s SEED, --seed SEED  the seed value to use for the random number generator; randomly seeded if
                        not provided (default: None)
  -a, --seed-augmentation
                        whether to seed the augmentation; if specified, uses the seeded random
                        generator to produce a seed value from 0 to 1000 for the augmentation.
                        (default: False)
  -T THRESHOLD, --threshold THRESHOLD
                        the threshold to use for Random.rand(): if equal or above, augmentation gets
                        applied; range: 0-1; default: 0 (= always) (default: None)

LABEL-PRESENT#

Keeps or discards images depending on whether annotations with certain label(s) are present. Checks can be further tightened by defining regions in the image that annotations must overlap with (or not overlap at all).

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: label-present [--coordinate-separator CHAR] [--invert-regions] [-l LABELS [LABELS ...]]
                     [--min-iou FLOAT] [--pair-separator CHAR] [-r regexp]
                     [--region [x,y[;x,y[;...]] [x,y[;x,y[;...]] ...]]] [--verbose]

optional arguments:
  --coordinate-separator CHAR
                        the separator between coordinates (default: ;)
  --invert-regions      Inverts the matching sense from 'labels have to overlap at least one of the
                        region(s)' to 'labels cannot overlap any region' (default: False)
  -l LABELS [LABELS ...], --labels LABELS [LABELS ...]
                        explicit list of labels to check (default: [])
  --min-iou FLOAT       the minimum IoU (intersect over union) that the object must have with the
                        region(s) in order to be considered an overlap (object detection only)
                        (default: 0.01)
  --pair-separator CHAR
                        the separator between the x and y of a pair (default: ,)
  -r regexp, --regexp regexp
                        regular expression for using only a subset of labels (default: None)
  --region [x,y[;x,y[;...]] [x,y[;x,y[;...]] ...]]
                        semicolon-separated list of comma-separated x/y pairs defining the region
                        that the object must overlap with in order to be included. Values between
                        0-1 are considered normalized, otherwise absolute pixels. (default: None)
  --verbose             Outputs some debugging information (default: False)

LINEAR-CONTRAST#

Applies linear contrast to images.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
Options:#
usage: linear-contrast [-f ALPHA_FROM] [-t ALPHA_TO] [-m IMGAUG_MODE] [--suffix IMGAUG_SUFFIX]
                       [-s SEED] [-a] [-T THRESHOLD]

optional arguments:
  -f ALPHA_FROM, --from-alpha ALPHA_FROM
                        the minimum alpha to apply to the images (default: None)
  -t ALPHA_TO, --to-alpha ALPHA_TO
                        the maximum alpha to apply to the images (default: None)
  -m IMGAUG_MODE, --mode IMGAUG_MODE
                        the image augmentation mode to use, available modes: replace, add (default:
                        replace)
  --suffix IMGAUG_SUFFIX
                        the suffix to use for the file names in case of augmentation mode add
                        (default: None)
  -s SEED, --seed SEED  the seed value to use for the random number generator; randomly seeded if
                        not provided (default: None)
  -a, --seed-augmentation
                        whether to seed the augmentation; if specified, uses the seeded random
                        generator to produce a seed value from 0 to 1000 for the augmentation.
                        (default: False)
  -T THRESHOLD, --threshold THRESHOLD
                        the threshold to use for Random.rand(): if equal or above, augmentation gets
                        applied; range: 0-1; default: 0 (= always) (default: None)

MAP-LABELS#

Maps object-detection labels from one set to another

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: map-labels [-m old=new]

optional arguments:
  -m old=new, --mapping old=new
                        mapping for labels, for replacing one label string with another (eg when
                        fixing/collapsing labels) (default: [])

MEL-SPECTROGRAM#

Generates a plot from a Mel spectrogram.

Domain(s):#
  • Audio classification domain
Options:#
usage: mel-spectrogram [--center] [--dpi DPI] [--hop-length HOP_LENGTH] [--num-fft NUM_FFT]
                       [--pad-mode PAD_MODE] [--power POWER] [--win-length WIN_LENGTH]
                       [--window WINDOW]

optional arguments:
  --center              for centering the signal. (default: False)
  --dpi DPI             the dots per inch (default: 100)
  --hop-length HOP_LENGTH
                        number of audio samples between adjacent STFT columns. (default: 512)
  --num-fft NUM_FFT     the length of the windowed signal after padding with zeros. should be power
                        of two. (default: 2048)
  --pad-mode PAD_MODE   used when 'centering' (default: constant)
  --power POWER         exponent for the magnitude melspectrogram. e.g., 1 for energy, 2 for power,
                        etc. (default: 2.0)
  --win-length WIN_LENGTH
                        each frame of audio is windowed by window of length win_length and then
                        padded with zeros to match num_fft. defaults to win_length = num_fft
                        (default: None)
  --window WINDOW       a window function, such as scipy.signal.windows.hann (default: hann)

MFCC-SPECTROGRAM#

Generates a plot from Mel-frequency cepstral coefficients.

Domain(s):#
  • Audio classification domain
Options:#
usage: mfcc-spectrogram [--center] [--dct-type DCT_TYPE] [--dpi DPI] [--hop-length HOP_LENGTH]
                        [--lifter LIFTER] [--norm NORM] [--num-fft NUM_FFT] [--num-mfcc NUM_MFCC]
                        [--pad-mode PAD_MODE] [--power POWER] [--win-length WIN_LENGTH]
                        [--window WINDOW]

optional arguments:
  --center              for centering the signal. (default: False)
  --dct-type DCT_TYPE   the Discrete cosine transform (DCT) type (1|2|3). By default, DCT type-2 is
                        used. (default: 2)
  --dpi DPI             the dots per inch (default: 100)
  --hop-length HOP_LENGTH
                        number of audio samples between adjacent STFT columns. (default: 512)
  --lifter LIFTER       If lifter>0, apply liftering (cepstral filtering) to the MFCC: M[n, :] <-
                        M[n, :] * (1 + sin(pi * (n + 1) / lifter) * lifter / 2) (default: 0)
  --norm NORM           If dct_type is 2 or 3, setting norm='ortho' uses an ortho-normal DCT basis.
                        Normalization is not supported for dct_type=1. (options: none|ortho)
                        (default: ortho)
  --num-fft NUM_FFT     the length of the windowed signal after padding with zeros. should be power
                        of two. (default: 2048)
  --num-mfcc NUM_MFCC   the number of MFCCs to return. (default: 20)
  --pad-mode PAD_MODE   used when 'centering' (default: constant)
  --power POWER         exponent for the magnitude melspectrogram. e.g., 1 for energy, 2 for power,
                        etc. (default: 2.0)
  --win-length WIN_LENGTH
                        each frame of audio is windowed by window of length win_length and then
                        padded with zeros to match num_fft. defaults to win_length = num_fft
                        (default: None)
  --window WINDOW       a window function, such as scipy.signal.windows.hann (default: hann)

OD-TO-IC#

Converts image object-detection instances into image classification instances

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: od-to-ic [-m HANDLER]

optional arguments:
  -m HANDLER, --multiplicity HANDLER
                        how to handle instances with more than one located object (default: error)

OD-TO-IS#

Converts image object-detection instances into image segmentation instances

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: od-to-is [--label-error] --labels LABEL [LABEL ...]

optional arguments:
  --label-error         whether to raise errors when an unspecified label is encountered (default is
                        to ignore) (default: False)
  --labels LABEL [LABEL ...]
                        specifies the labels for each index (default: None)

PASSTHROUGH#

Dummy ISP which has no effect on the conversion stream

Domain(s):#
  • Speech Domain
  • Image Object-Detection Domain
  • Image Classification Domain
  • Image Segmentation Domain
  • Audio classification domain
Options:#
usage: passthrough

PITCH-SHIFT#

Augmentation method for shifting the pitch of audio files.

Domain(s):#
  • Speech Domain
  • Audio classification domain
Options:#
usage: pitch-shift [-m AUG_MODE] [--suffix AUG_SUFFIX] [--bins-per-octave BINS_PER_OCTAVE]
                   [--resample-type RESAMPLE_TYPE] [-s SEED] [-a] [-f STEPS_FROM] [-t STEPS_TO]
                   [-T THRESHOLD] [-v]

optional arguments:
  -m AUG_MODE, --mode AUG_MODE
                        the audio augmentation mode to use, available modes: replace, add (default:
                        replace)
  --suffix AUG_SUFFIX   the suffix to use for the file names in case of augmentation mode add
                        (default: None)
  --bins-per-octave BINS_PER_OCTAVE
                        how many steps per octave (default: 12)
  --resample-type RESAMPLE_TYPE
                        the resampling type to apply (kaiser_best|kaiser_fast|fft|polyphase|linear|z
                        ero_order_hold|sinc_best|sinc_medium|sinc_fastest|soxr_vhq|soxr_hq|soxr_mq|s
                        oxr_lq|soxr_qq) (default: kaiser_best)
  -s SEED, --seed SEED  the seed value to use for the random number generator; randomly seeded if
                        not provided (default: None)
  -a, --seed-augmentation
                        whether to seed the augmentation; if specified, uses the seeded random
                        generator to produce a seed value from 0 to 1000 for the augmentation.
                        (default: False)
  -f STEPS_FROM, --from-steps STEPS_FROM
                        the minimum (fractional) steps to shift (default: None)
  -t STEPS_TO, --to-steps STEPS_TO
                        the maximum (fractional) steps to shift (default: None)
  -T THRESHOLD, --threshold THRESHOLD
                        the threshold to use for Random.rand(): if equal or above, augmentation gets
                        applied; range: 0-1; default: 0 (= always) (default: None)
  -v, --verbose         whether to output debugging information (default: False)

POLYGON-DISCARDER#

Removes annotations with polygons which fall outside certain point limit constraints

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: polygon-discarder [--max-points MAX_POINTS] [--min-points MIN_POINTS] [--verbose]

optional arguments:
  --max-points MAX_POINTS
                        the maximum number of points in the polygon (default: None)
  --min-points MIN_POINTS
                        the minimum number of points in the polygon (default: None)
  --verbose             outputs information when discarding annotations (default: False)

REDIS-PREDICT-IC#

Makes image classification predictions via Redis backend, passing in an image and receiving JSON predictions back (at least one of 'label: probability'). Predictions example:

Domain(s):#
  • Image Classification Domain
Options:#
usage: redis-predict-ic [--channel-in CHANNEL_IN] [--channel-out CHANNEL_OUT] [-d REDIS_DB]
                        [-h REDIS_HOST] [-p REDIS_PORT] [-t TIMEOUT] [-v]

optional arguments:
  --channel-in CHANNEL_IN
                        the Redis channel on which to receive predictions. (default: predictions)
  --channel-out CHANNEL_OUT
                        the Redis channel to send the images out (default: images)
  -d REDIS_DB, --redis-db REDIS_DB
                        the database to use (default: 0)
  -h REDIS_HOST, --redis-host REDIS_HOST
                        the Redis server to connect to (default: localhost)
  -p REDIS_PORT, --redis-port REDIS_PORT
                        the port the Redis server is running on (default: 6379)
  -t TIMEOUT, --timeout TIMEOUT
                        the timeout in seconds to wait for a prediction to arrive (default: 5.0)
  -v, --verbose         whether to output debugging information. (default: False)

REDIS-PREDICT-IS#

Makes image segmentation predictions via Redis backend, passing in an image and receiving an image with predicted segmentations.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: redis-predict-is [--channel-in CHANNEL_IN] [--channel-out CHANNEL_OUT]
                        [--image-format IMAGE_FORMAT] --labels LABEL [LABEL ...] [-d REDIS_DB]
                        [-h REDIS_HOST] [-p REDIS_PORT] [-t TIMEOUT] [-v]

optional arguments:
  --channel-in CHANNEL_IN
                        the Redis channel on which to receive predictions. (default: predictions)
  --channel-out CHANNEL_OUT
                        the Redis channel to send the images out (default: images)
  --image-format IMAGE_FORMAT
                        the format of the image that comes back as prediction:
                        indexedpng,bluechannel,grayscale (default: indexedpng)
  --labels LABEL [LABEL ...]
                        specifies the labels for each index (default: None)
  -d REDIS_DB, --redis-db REDIS_DB
                        the database to use (default: 0)
  -h REDIS_HOST, --redis-host REDIS_HOST
                        the Redis server to connect to (default: localhost)
  -p REDIS_PORT, --redis-port REDIS_PORT
                        the port the Redis server is running on (default: 6379)
  -t TIMEOUT, --timeout TIMEOUT
                        the timeout in seconds to wait for a prediction to arrive (default: 5.0)
  -v, --verbose         whether to output debugging information. (default: False)

REDIS-PREDICT-OD#

Makes object detection predictions via Redis backend, passing in an image and receiving OPEX predictions back: https://github.com/WaikatoLink2020/objdet-predictions-exchange-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: redis-predict-od [--channel-in CHANNEL_IN] [--channel-out CHANNEL_OUT]
                        [--key-label KEY_LABEL] [--key-score KEY_SCORE] [-d REDIS_DB]
                        [-h REDIS_HOST] [-p REDIS_PORT] [-t TIMEOUT] [-v]

optional arguments:
  --channel-in CHANNEL_IN
                        the Redis channel on which to receive predictions. (default: predictions)
  --channel-out CHANNEL_OUT
                        the Redis channel to send the images out (default: images)
  --key-label KEY_LABEL
                        the meta-data key in the annotations to use for storing the label. (default:
                        type)
  --key-score KEY_SCORE
                        the meta-data key in the annotations to use for storing the prediction
                        score. (default: score)
  -d REDIS_DB, --redis-db REDIS_DB
                        the database to use (default: 0)
  -h REDIS_HOST, --redis-host REDIS_HOST
                        the Redis server to connect to (default: localhost)
  -p REDIS_PORT, --redis-port REDIS_PORT
                        the port the Redis server is running on (default: 6379)
  -t TIMEOUT, --timeout TIMEOUT
                        the timeout in seconds to wait for a prediction to arrive (default: 5.0)
  -v, --verbose         whether to output debugging information. (default: False)

REMOVE-CLASSES#

Removes classes from classification/image-segmentation instances

Domain(s):#
  • Image Classification Domain
  • Image Segmentation Domain
  • Audio classification domain
Options:#
usage: remove-classes -c CLASS [CLASS ...]

optional arguments:
  -c CLASS [CLASS ...], --classes CLASS [CLASS ...]
                        the classes to remove (default: None)

RENAME#

ISP that renames files.

Domain(s):#
  • Speech Domain
  • Image Object-Detection Domain
  • Image Classification Domain
  • Image Segmentation Domain
  • Audio classification domain
Options:#
usage: rename [-f NAME_FORMAT] [--verbose]

optional arguments:
  -f NAME_FORMAT, --name-format NAME_FORMAT
                        the format for the new name. Available placeholders: - {name}: the name of
                        the file, without path or extension. - {ext}: the extension of the file
                        (incl dot). - {occurrences}: the number of times this name (excl extension)
                        has been encountered. - {count}: the number of files encountered so far. -
                        {[p]+dir}: the parent directory of the file: 'p': immediate parent, the more
                        the p's the higher up in the hierarchy. (default: {name}{ext})
  --verbose             outputs information about generated names (default: False)

RESAMPLE-AUDIO#

Resamples audio files.

For resample types, see: https://librosa.org/doc/latest/generated/librosa.resample.html#librosa.resample

Domain(s):#
  • Speech Domain
  • Audio classification domain
Options:#
usage: resample-audio [-t RESAMPLE_TYPE] [-s SAMPLE_RATE] [-v]

optional arguments:
  -t RESAMPLE_TYPE, --resample-type RESAMPLE_TYPE
                        the resampling type to apply (kaiser_best|kaiser_fast|fft|polyphase|linear|z
                        ero_order_hold|sinc_best|sinc_medium|sinc_fastest|soxr_vhq|soxr_hq|soxr_mq|s
                        oxr_lq|soxr_qq) (default: kaiser_best)
  -s SAMPLE_RATE, --sample-rate SAMPLE_RATE
                        the sample rate to use for the audio data. (default: 22050)
  -v, --verbose         whether to output some debugging output (default: False)

ROTATE#

Rotates images randomly within a range of degrees or by a specified degree. Specify seed value and force augmentation to be seeded to generate repeatable augmentations.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
Options:#
usage: rotate [-f DEGREE_FROM] [-t DEGREE_TO] [-m IMGAUG_MODE] [--suffix IMGAUG_SUFFIX] [-s SEED]
              [-a] [-T THRESHOLD]

optional arguments:
  -f DEGREE_FROM, --from-degree DEGREE_FROM
                        the start of the degree range to use for rotating the images (default: None)
  -t DEGREE_TO, --to-degree DEGREE_TO
                        the end of the degree range to use for rotating the images (default: None)
  -m IMGAUG_MODE, --mode IMGAUG_MODE
                        the image augmentation mode to use, available modes: replace, add (default:
                        replace)
  --suffix IMGAUG_SUFFIX
                        the suffix to use for the file names in case of augmentation mode add
                        (default: None)
  -s SEED, --seed SEED  the seed value to use for the random number generator; randomly seeded if
                        not provided (default: None)
  -a, --seed-augmentation
                        whether to seed the augmentation; if specified, uses the seeded random
                        generator to produce a seed value from 0 to 1000 for the augmentation.
                        (default: False)
  -T THRESHOLD, --threshold THRESHOLD
                        the threshold to use for Random.rand(): if equal or above, augmentation gets
                        applied; range: 0-1; default: 0 (= always) (default: None)

SAMPLE#

ISP that selects a subset from the stream.

Domain(s):#
  • Speech Domain
  • Image Object-Detection Domain
  • Image Classification Domain
  • Image Segmentation Domain
  • Audio classification domain
Options:#
usage: sample [-s SEED] [-T THRESHOLD]

optional arguments:
  -s SEED, --seed SEED  the seed value to use for the random number generator; randomly seeded if
                        not provided (default: None)
  -T THRESHOLD, --threshold THRESHOLD
                        the threshold to use for Random.rand(): if equal or above, sample gets
                        selected; range: 0-1; default: 0 (= always) (default: 0.0)

SCALE#

Scales images randomly within a range of percentages or by a specified percentage. Specify seed value and force augmentation to be seeded to generate repeatable augmentations.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
Options:#
usage: scale [-m IMGAUG_MODE] [--suffix IMGAUG_SUFFIX] [-k] [-f PERCENTAGE_FROM] [-t PERCENTAGE_TO]
             [-s SEED] [-a] [-T THRESHOLD] [-u]

optional arguments:
  -m IMGAUG_MODE, --mode IMGAUG_MODE
                        the image augmentation mode to use, available modes: replace, add (default:
                        replace)
  --suffix IMGAUG_SUFFIX
                        the suffix to use for the file names in case of augmentation mode add
                        (default: None)
  -k, --keep-aspect     whether to keep the aspect ratio (default: False)
  -f PERCENTAGE_FROM, --from-percentage PERCENTAGE_FROM
                        the start of the percentage range to use for scaling the images (default:
                        None)
  -t PERCENTAGE_TO, --to-percentage PERCENTAGE_TO
                        the end of the percentage range to use for scaling the images (default:
                        None)
  -s SEED, --seed SEED  the seed value to use for the random number generator; randomly seeded if
                        not provided (default: None)
  -a, --seed-augmentation
                        whether to seed the augmentation; if specified, uses the seeded random
                        generator to produce a seed value from 0 to 1000 for the augmentation.
                        (default: False)
  -T THRESHOLD, --threshold THRESHOLD
                        the threshold to use for Random.rand(): if equal or above, augmentation gets
                        applied; range: 0-1; default: 0 (= always) (default: None)
  -u, --update-size     whether to update the image size after the scaling operation or use original
                        size (default: False)

SKIP-SIMILAR-FRAMES#

Skips frames in the stream that are deemed too similar.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
  • Image Segmentation Domain
Options:#
usage: skip-similar-frames [-b BW_THRESHOLD] [-t CHANGE_THRESHOLD] [-c CONVERSION] [-v]

optional arguments:
  -b BW_THRESHOLD, --bw-threshold BW_THRESHOLD
                        the threshold to use for converting a gray-scale like image to black and
                        white (0-255) (default: 128)
  -t CHANGE_THRESHOLD, --change-threshold CHANGE_THRESHOLD
                        the percentage of pixels that changed relative to size of image (0-1)
                        (default: 0.01)
  -c CONVERSION, --conversion CONVERSION
                        how to convert the BGR image to a single channel image (gray/r/g/b)
                        (default: gray)
  -v, --verbose         whether to output some debugging output. (default: False)

STFT-SPECTROGRAM#

Generates a plot from a short time fourier transform (STFT) spectrogram.

Domain(s):#
  • Audio classification domain
Options:#
usage: stft-spectrogram [--center] [--dpi DPI] [--hop-length HOP_LENGTH] [--num-fft NUM_FFT]
                        [--pad-mode PAD_MODE] [--win-length WIN_LENGTH] [--window WINDOW]

optional arguments:
  --center              for centering the signal. (default: False)
  --dpi DPI             the dots per inch (default: 100)
  --hop-length HOP_LENGTH
                        number of audio samples between adjacent STFT columns. defaults to
                        win_length // 4 (default: None)
  --num-fft NUM_FFT     the length of the windowed signal after padding with zeros. should be power
                        of two. (default: 2048)
  --pad-mode PAD_MODE   used when 'centering' (default: constant)
  --win-length WIN_LENGTH
                        each frame of audio is windowed by window of length win_length and then
                        padded with zeros to match num_fft. defaults to win_length = num_fft
                        (default: None)
  --window WINDOW       a window function, such as scipy.signal.windows.hann (default: hann)

STRIP-ANNOTATIONS#

ISP which removes annotations from instances

Domain(s):#
  • Speech Domain
  • Image Object-Detection Domain
  • Image Classification Domain
  • Image Segmentation Domain
  • Audio classification domain
Options:#
usage: strip-annotations

SUB-IMAGES#

Extracts sub-images (incl their annotations) from the images coming through, using the defined regions.

Domain(s):#
  • Image Classification Domain
  • Image Object-Detection Domain
Options:#
usage: sub-images [-p] [-s REGION_SORTING] [-r REGIONS [REGIONS ...]] [-e] [--verbose]

optional arguments:
  -p, --include-partial
                        whether to include only annotations that fit fully into a region or also
                        partial ones (default: False)
  -s REGION_SORTING, --region-sorting REGION_SORTING
                        how to sort the supplied region definitions: none|x-then-y|y-then-x
                        (default: none)
  -r REGIONS [REGIONS ...], --regions REGIONS [REGIONS ...]
                        the regions (X,Y,WIDTH,HEIGHT) to crop and forward with their annotations
                        (default: [])
  -e, --suppress-empty  suppresses sub-images that have no annotations (object detection) (default:
                        False)
  --verbose             for outputting debugging information (default: False)

TIME-STRETCH#

Augmentation method for stretching the time of audio files (speed up/slow down).

Domain(s):#
  • Speech Domain
  • Audio classification domain
Options:#
usage: time-stretch [-m AUG_MODE] [--suffix AUG_SUFFIX] [-f RATE_FROM] [-t RATE_TO] [-s SEED] [-a]
                    [-T THRESHOLD] [-v]

optional arguments:
  -m AUG_MODE, --mode AUG_MODE
                        the audio augmentation mode to use, available modes: replace, add (default:
                        replace)
  --suffix AUG_SUFFIX   the suffix to use for the file names in case of augmentation mode add
                        (default: None)
  -f RATE_FROM, --from-rate RATE_FROM
                        the minimum stretch factor (<1: slow down, 1: same, >1: speed up) (default:
                        None)
  -t RATE_TO, --to-rate RATE_TO
                        the maximum stretch factor (<1: slow down, 1: same, >1: speed up) (default:
                        None)
  -s SEED, --seed SEED  the seed value to use for the random number generator; randomly seeded if
                        not provided (default: None)
  -a, --seed-augmentation
                        whether to seed the augmentation; if specified, uses the seeded random
                        generator to produce a seed value from 0 to 1000 for the augmentation.
                        (default: False)
  -T THRESHOLD, --threshold THRESHOLD
                        the threshold to use for Random.rand(): if equal or above, augmentation gets
                        applied; range: 0-1; default: 0 (= always) (default: None)
  -v, --verbose         whether to output debugging information (default: False)

TRIM-AUDIO#

Trims silence from audio files.

Domain(s):#
  • Speech Domain
  • Audio classification domain
Options:#
usage: trim-audio [--frame-length FRAME_LENGTH] [--hop-length HOP_LENGTH] [--top-db TOP_DB] [-v]

optional arguments:
  --frame-length FRAME_LENGTH
                        the number of samples per analysis frame. (default: 2048)
  --hop-length HOP_LENGTH
                        the number of samples between analysis frames (default: 512)
  --top-db TOP_DB       the threshold (in decibels) below reference to consider as silence.
                        (default: 60)
  -v, --verbose         whether to output some debugging output (default: False)

Sink stage#

AREA-HISTOGRAM-IS#

Generates histograms of the area (normalized or absolute) occupied by the annotations.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: area-histogram-is [-a ALL_LABEL] [-b] [--label-key LABEL_KEY] [-n] [--num-bins NUM_BINS]
                         [-o OUTPUT_FILE] [-f OUTPUT_FORMAT]

optional arguments:
  -a ALL_LABEL, --all-label ALL_LABEL
                        the label to use for all the labels combined (default: ALL)
  -b, --force-bbox      whether to use the bounding box even if a polygon is present (object
                        detection domain only) (default: False)
  --label-key LABEL_KEY
                        the key in the meta-data that contains the label. (default: type)
  -n, --normalized      whether to use normalized areas (using the image size as base). (default:
                        False)
  --num-bins NUM_BINS   the number of bins to use for the histogram. (default: 20)
  -o OUTPUT_FILE, --output OUTPUT_FILE
                        the file to write the histogram to; uses stdout if omitted (default: )
  -f OUTPUT_FORMAT, --format OUTPUT_FORMAT
                        the format to use for the output, available modes: csv, json (default: text)

AREA-HISTOGRAM-OD#

Generates histograms of the area (normalized or absolute) occupied by the annotations.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: area-histogram-od [-a ALL_LABEL] [-b] [--label-key LABEL_KEY] [-n] [--num-bins NUM_BINS]
                         [-o OUTPUT_FILE] [-f OUTPUT_FORMAT]

optional arguments:
  -a ALL_LABEL, --all-label ALL_LABEL
                        the label to use for all the labels combined (default: ALL)
  -b, --force-bbox      whether to use the bounding box even if a polygon is present (object
                        detection domain only) (default: False)
  --label-key LABEL_KEY
                        the key in the meta-data that contains the label. (default: type)
  -n, --normalized      whether to use normalized areas (using the image size as base). (default:
                        False)
  --num-bins NUM_BINS   the number of bins to use for the histogram. (default: 20)
  -o OUTPUT_FILE, --output OUTPUT_FILE
                        the file to write the histogram to; uses stdout if omitted (default: )
  -f OUTPUT_FORMAT, --format OUTPUT_FORMAT
                        the format to use for the output, available modes: csv, json (default: text)

AUDIO-INFO-AC#

Collates and outputs information on the audio files.

Domain(s):#
  • Audio classification domain
Options:#
usage: audio-info-ac [-o OUTPUT_FILE] [-f OUTPUT_FORMAT]

optional arguments:
  -o OUTPUT_FILE, --output OUTPUT_FILE
                        the file to write the information to; uses stdout if omitted (default: )
  -f OUTPUT_FORMAT, --format OUTPUT_FORMAT
                        the format to use for the output, available modes: csv, json (default: text)

AUDIO-INFO-SP#

Collates and outputs information on the audio files.

Domain(s):#
  • Speech Domain
Options:#
usage: audio-info-sp [-o OUTPUT_FILE] [-f OUTPUT_FORMAT]

optional arguments:
  -o OUTPUT_FILE, --output OUTPUT_FILE
                        the file to write the information to; uses stdout if omitted (default: )
  -f OUTPUT_FORMAT, --format OUTPUT_FORMAT
                        the format to use for the output, available modes: csv, json (default: text)

CALC-FRAME-CHANGES#

Calculates the changes between frames, which can be used with the skip-similar-frames ISP.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: calc-frame-changes [-b BW_THRESHOLD] [-t CHANGE_THRESHOLD] [-c CONVERSION] [-B NUM_BINS]
                          [-o OUTPUT_FILE] [-f OUTPUT_FORMAT] [-v]

optional arguments:
  -b BW_THRESHOLD, --bw-threshold BW_THRESHOLD
                        the threshold to use for converting a gray-scale like image to black and
                        white (0-255) (default: 128)
  -t CHANGE_THRESHOLD, --change-threshold CHANGE_THRESHOLD
                        the percentage of pixels that changed relative to size of image (0-1)
                        (default: 0.01)
  -c CONVERSION, --conversion CONVERSION
                        how to convert the BGR image to a single channel image (gray/r/g/b)
                        (default: gray)
  -B NUM_BINS, --num-bins NUM_BINS
                        the number of bins to use for the histogram (default: 20)
  -o OUTPUT_FILE, --output OUTPUT_FILE
                        the file to write to statistics to, stdout if not provided (default: )
  -f OUTPUT_FORMAT, --output-format OUTPUT_FORMAT
                        how to output the statistics (text/csv/json) (default: text)
  -v, --verbose         whether to output some debugging output. (default: False)

GENERIC-SINK-AC#

Generic audio classification sink.

Domain(s):#
  • Audio classification domain
Options:#
usage: generic-sink-ac [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-SINK-IC#

Generic image classification sink.

Domain(s):#
  • Image Classification Domain
Options:#
usage: generic-sink-ic [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-SINK-IS#

Generic image segmentation sink.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: generic-sink-is [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-SINK-OD#

Generic object detection sink.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: generic-sink-od [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

GENERIC-SINK-SP#

Generic speech sink.

Domain(s):#
  • Speech Domain
Options:#
usage: generic-sink-sp [-c USER_CLASS] [-o USER_OPTIONS]

optional arguments:
  -c USER_CLASS, --class USER_CLASS
                        the user class to wrap (dot notation) (default: None)
  -o USER_OPTIONS, --options USER_OPTIONS
                        the options for the user class to parse (default: None)

IMAGE-VIEWER-IC#

Displays image classification images.

Domain(s):#
  • Image Classification Domain
Options:#
usage: image-viewer-ic [--delay DELAY] [--position POSITION] [--size SIZE] [--title TITLE]

optional arguments:
  --delay DELAY        the delay in milli-seconds between images, use 0 to wait for keypress,
                       ignored if <0 (default: 500)
  --position POSITION  the position of the window on screen (X,Y) (default: 0,0)
  --size SIZE          the maximum size for the image: WIDTH,HEIGHT (default: 640,480)
  --title TITLE        the title for the window (default: wai.annotations)

IMAGE-VIEWER-IS#

Displays image segmentation images.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: image-viewer-is [--delay DELAY] [--position POSITION] [--size SIZE] [--title TITLE]

optional arguments:
  --delay DELAY        the delay in milli-seconds between images, use 0 to wait for keypress,
                       ignored if <0 (default: 500)
  --position POSITION  the position of the window on screen (X,Y) (default: 0,0)
  --size SIZE          the maximum size for the image: WIDTH,HEIGHT (default: 640,480)
  --title TITLE        the title for the window (default: wai.annotations)

IMAGE-VIEWER-OD#

Displays object detection images.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: image-viewer-od [--delay DELAY] [--position POSITION] [--size SIZE] [--title TITLE]

optional arguments:
  --delay DELAY        the delay in milli-seconds between images, use 0 to wait for keypress,
                       ignored if <0 (default: 500)
  --position POSITION  the position of the window on screen (X,Y) (default: 0,0)
  --size SIZE          the maximum size for the image: WIDTH,HEIGHT (default: 640,480)
  --title TITLE        the title for the window (default: wai.annotations)

LABEL-DIST-IC#

Generates a label distribution.

Domain(s):#
  • Image Classification Domain
Options:#
usage: label-dist-ic [--label-key LABEL_KEY] [-o OUTPUT_FILE] [-f OUTPUT_FORMAT] [-p]

optional arguments:
  --label-key LABEL_KEY
                        the key in the meta-data that contains the label. (default: type)
  -o OUTPUT_FILE, --output OUTPUT_FILE
                        the file to write the statistics to; uses stdout if omitted (default: )
  -f OUTPUT_FORMAT, --format OUTPUT_FORMAT
                        the format to use for the output, available modes: csv, json (default: text)
  -p, --percentages     whether to output percentages instead of counts. (default: False)

LABEL-DIST-IS#

Generates a label distribution.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: label-dist-is [--label-key LABEL_KEY] [-o OUTPUT_FILE] [-f OUTPUT_FORMAT] [-p]

optional arguments:
  --label-key LABEL_KEY
                        the key in the meta-data that contains the label. (default: type)
  -o OUTPUT_FILE, --output OUTPUT_FILE
                        the file to write the statistics to; uses stdout if omitted (default: )
  -f OUTPUT_FORMAT, --format OUTPUT_FORMAT
                        the format to use for the output, available modes: csv, json (default: text)
  -p, --percentages     whether to output percentages instead of counts. (default: False)

LABEL-DIST-OD#

Generates a label distribution.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: label-dist-od [--label-key LABEL_KEY] [-o OUTPUT_FILE] [-f OUTPUT_FORMAT] [-p]

optional arguments:
  --label-key LABEL_KEY
                        the key in the meta-data that contains the label. (default: type)
  -o OUTPUT_FILE, --output OUTPUT_FILE
                        the file to write the statistics to; uses stdout if omitted (default: )
  -f OUTPUT_FORMAT, --format OUTPUT_FORMAT
                        the format to use for the output, available modes: csv, json (default: text)
  -p, --percentages     whether to output percentages instead of counts. (default: False)

TO-ADAMS-IC#

Writes image classification annotations in the ADAMS report-format

Domain(s):#
  • Image Classification Domain
Options:#
usage: to-adams-ic -c FIELD [--annotations-only] [--no-interleave] -o PATH
                   [--split-names SPLIT NAME [SPLIT NAME ...]] [--split-ratios RATIO [RATIO ...]]

optional arguments:
  -c FIELD, --class-field FIELD
                        the report field containing the image class (default: None)
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        output directory to write files to (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-ADAMS-OD#

Writes image object-detection annotations in the ADAMS report-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-adams-od [--annotations-only] [--no-interleave] -o PATH
                   [--split-names SPLIT NAME [SPLIT NAME ...]] [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        output directory to write files to (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-ANNOTATION-OVERLAY-OD#

Generates an image with all the annotation shapes (bbox or polygon) overlayed.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-annotation-overlay-od [-b BACKGROUND_COLOR] [-c COLOR] [-o OUTPUT_FILE] [-s SCALE_TO]

optional arguments:
  -b BACKGROUND_COLOR, --background-color BACKGROUND_COLOR
                        the color to use for the background as RGBA byte-quadruplet, e.g.:
                        255,255,255,255 (default: 255,255,255,255)
  -c COLOR, --color COLOR
                        the color to use for drawing the shapes as RGBA byte-quadruplet, e.g.:
                        255,0,0,64 (default: 255,0,0,64)
  -o OUTPUT_FILE, --output OUTPUT_FILE
                        the PNG image to write the generated overlay to (default: ./overlay.png)
  -s SCALE_TO, --scale-to SCALE_TO
                        the dimensions to scale all images to before overlaying them (format:
                        width,height) (default: )

TO-AUDIO-FILES-AC#

Dummy writer that just outputs audio files from classification datasets.

Domain(s):#
  • Audio classification domain
Options:#
usage: to-audio-files-ac [-o OUTPUT_DIR]

optional arguments:
  -o OUTPUT_DIR, --output-dir OUTPUT_DIR
                        the directory to write the audio files to (default: .)

TO-AUDIO-FILES-SP#

Dummy writer that just outputs audio files from speech datasets.

Domain(s):#
  • Speech Domain
Options:#
usage: to-audio-files-sp [-o OUTPUT_DIR]

optional arguments:
  -o OUTPUT_DIR, --output-dir OUTPUT_DIR
                        the directory to write the audio files to (default: .)

TO-BLUE-CHANNEL-IS#

Writes image segmentation files in the blue-channel format

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: to-blue-channel-is [--annotations-only] [--no-interleave] -o PATH
                          [--split-names SPLIT NAME [SPLIT NAME ...]]
                          [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        the directory to write the annotation images to (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-COCO-OD#

Writes image object-detection annotations in the MS-COCO JSON-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-coco-od [--annotations-only] [--categories CATEGORY [CATEGORY ...]]
                  [--category-output-file FILENAME] [--default-supercategory SUPERCATEGORY]
                  [--error-on-new-category] [--license-name LICENSE_NAME]
                  [--license-url LICENSE_URL] [--no-interleave] -o PATH [--pretty]
                  [--sort-categories] [--split-names SPLIT NAME [SPLIT NAME ...]]
                  [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --categories CATEGORY [CATEGORY ...]
                        defines the order of the categories (default: [])
  --category-output-file FILENAME
                        file to write the categories into, as a simple comma-separated list
                        (default: None)
  --default-supercategory SUPERCATEGORY
                        the supercategory to use for pre-defined categories (default: Object)
  --error-on-new-category
                        whether unspecified categories should raise an error (default: False)
  --license-name LICENSE_NAME
                        the license of the images (default: default)
  --license-url LICENSE_URL
                        the license of the images (default: )
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        output file to write annotations to (images are placed in same directory)
                        (default: None)
  --pretty              whether to format the JSON annotations file with indentation (default:
                        False)
  --sort-categories     whether to put the categories in alphabetical order (default: False)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-COMMON-VOICE-SP#

Writes speech transcriptions in the Mozilla Common-Voice TSV-format

Domain(s):#
  • Speech Domain
Options:#
usage: to-common-voice-sp [--annotations-only] [--no-interleave] -o PATH
                          [--split-names SPLIT NAME [SPLIT NAME ...]]
                          [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        the filename of the TSV file to write the annotations into (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-COQUI-STT-SP#

Writes speech transcriptions in the Coqui STT CSV-format

Domain(s):#
  • Speech Domain
Options:#
usage: to-coqui-stt-sp [--annotations-only] [--no-interleave] -o PATH
                       [--split-names SPLIT NAME [SPLIT NAME ...]]
                       [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        the filename of the CSV file to write the annotations into (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-COQUI-TTS-SP#

Writes speech transcriptions in the Coqui TTS text-format

Domain(s):#
  • Speech Domain
Options:#
usage: to-coqui-tts-sp [--annotations-only] [--no-interleave] -o PATH
                       [--split-names SPLIT NAME [SPLIT NAME ...]]
                       [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        the filename of the TTS file to write the annotations into (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-FESTVOX-SP#

Writes speech transcriptions in the Festival FestVox format

Domain(s):#
  • Speech Domain
Options:#
usage: to-festvox-sp [--annotations-only] [--no-interleave] -o PATH
                     [--split-names SPLIT NAME [SPLIT NAME ...]] [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        the filename of the FestVox file to write the annotations into (default:
                        None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-GRAYSCALE-IS#

Writes image segmentation files in the grayscale format

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: to-grayscale-is [--annotations-only] [--no-interleave] -o PATH
                       [--split-names SPLIT NAME [SPLIT NAME ...]]
                       [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        the directory to write the annotation images to (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-IMAGES-IC#

Dummy writer that just outputs images from image classification datasets.

Domain(s):#
  • Image Classification Domain
Options:#
usage: to-images-ic [-o OUTPUT_DIR]

optional arguments:
  -o OUTPUT_DIR, --output-dir OUTPUT_DIR
                        the directory to write the images to (default: .)

TO-IMAGES-IS#

Dummy writer that just outputs images from image segmentation datasets.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: to-images-is [-o OUTPUT_DIR]

optional arguments:
  -o OUTPUT_DIR, --output-dir OUTPUT_DIR
                        the directory to write the images to (default: .)

TO-IMAGES-OD#

Dummy writer that just outputs images from object detection datasets.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-images-od [-o OUTPUT_DIR]

optional arguments:
  -o OUTPUT_DIR, --output-dir OUTPUT_DIR
                        the directory to write the images to (default: .)

TO-INDEXED-PNG-IS#

Writes image segmentation files in the indexed-PNG format

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: to-indexed-png-is [--annotations-only] [--no-interleave] -o PATH
                         [--split-names SPLIT NAME [SPLIT NAME ...]]
                         [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        the directory to write the annotation images to (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-LAYER-SEGMENTS-IS#

Writes the layer-segments image-segmentation format to disk

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: to-layer-segments-is [--annotations-only] [--label-separator SEPARATOR] [--no-interleave] -o
                            PATH [--split-names SPLIT NAME [SPLIT NAME ...]]
                            [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --label-separator SEPARATOR
                        the separator between the base filename and the label (default: -)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        the directory to write the annotation images to (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-OPEX-OD#

Writes image object-detection annotations in the OPEX format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-opex-od [-c PATH] [-l PATH] [--annotations-only] [--no-interleave] -o PATH
                  [--split-names SPLIT NAME [SPLIT NAME ...]] [--split-ratios RATIO [RATIO ...]]

optional arguments:
  -c PATH, --labels-csv PATH
                        Path to the labels CSV file to write (default: None)
  -l PATH, --labels PATH
                        Path to the labels file to write (default: None)
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        output directory to write images and annotations to (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-ROI-OD#

Writes image object-detection annotations in the ROI CSV-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-roi-od [-d WIDTH HEIGHT] [--annotations-only] [--comments COMMENTS [COMMENTS ...]]
                 [--no-interleave] -o PATH [--size-mode] [--split-names SPLIT NAME [SPLIT NAME ...]]
                 [--split-ratios RATIO [RATIO ...]] [--prefix WRITER_PREFIX]
                 [--suffix WRITER_SUFFIX]

optional arguments:
  -d WIDTH HEIGHT, --image-dimensions WIDTH HEIGHT
                        image dimensions to use if none can be inferred (default: [])
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --comments COMMENTS [COMMENTS ...]
                        comments to write to the beginning of the ROI file (default: [])
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        output directory to write files to (default: None)
  --size-mode           writes the ROI files with x,y,w,h headers instead of x0,y0,x1,y1 (default:
                        False)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])
  --prefix WRITER_PREFIX
                        the prefix for output filenames (default = '') (default: None)
  --suffix WRITER_SUFFIX
                        the suffix for output filenames (default = '-rois.csv') (default: None)

TO-SUBDIR-AC#

Writes audio files to sub-directories named after their class labels.

Domain(s):#
  • Audio classification domain
Options:#
usage: to-subdir-ac [--no-interleave] -o PATH [--split-names SPLIT NAME [SPLIT NAME ...]]
                    [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        the directory to store the class directories in (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-SUBDIR-IC#

Writes images to sub-directories named after their class labels.

Domain(s):#
  • Image Classification Domain
Options:#
usage: to-subdir-ic [--no-interleave] -o PATH [--split-names SPLIT NAME [SPLIT NAME ...]]
                    [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        the directory to store the class directories in (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-TF-OD#

Writes image object-detection annotations in the TFRecords binary format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-tf-od [--dense] [--source-id-type {filename,numeric-dummy}] [--no-interleave] -o PATH
                [-p FILENAME] [-s FILENAME [FILENAME ...]]
                [--split-names SPLIT NAME [SPLIT NAME ...]] [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --dense               outputs masks in the dense numerical format instead of PNG-encoded (default:
                        False)
  --source-id-type {filename,numeric-dummy}
                        by default, the filename gets stored in the 'source_id' field, but some
                        algorithms try to convert it into a number and fail with 'StringToNumberOp
                        could not correctly convert string'; in which case you can use 'numeric-
                        dummy' (see https://github.com/google/automl/issues/307) (default: filename)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        name of output file for TFRecords (default: None)
  -p FILENAME, --protobuf FILENAME
                        for storing the label strings and IDs (default: None)
  -s FILENAME [FILENAME ...], --shards FILENAME [FILENAME ...]
                        additional shards to write to (default: [])
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-VGG-OD#

Writes image object-detection annotations in the VGG JSON-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-vgg-od [--annotations-only] [--no-interleave] -o PATH [--pretty]
                 [--split-names SPLIT NAME [SPLIT NAME ...]] [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        output file to write annotations to (images are placed in same directory)
                        (default: None)
  --pretty              whether to format the JSON annotations file with indentation (default:
                        False)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-VIDEO-FILE-OD#

Writes frames to a MJPG video file.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-video-file-od [-f FPS] [-o OUTPUT_FILE]

optional arguments:
  -f FPS, --fps FPS     the frames per second to use (default: 25)
  -o OUTPUT_FILE, --output OUTPUT_FILE
                        the MJPG video file to write to (default: )

TO-VOC-OD#

Writes image object-detection annotations in the Pascal VOC XML-format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-voc-od [--annotations-only] [--no-interleave] -o PATH
                 [--split-names SPLIT NAME [SPLIT NAME ...]] [--split-ratios RATIO [RATIO ...]]

optional arguments:
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        output directory to write annotations to (images are placed in same
                        directory) (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])

TO-VOID-AC#

Consumes audio classification instances without writing them.

Domain(s):#
  • Audio classification domain
Options:#
usage: to-void-ac

TO-VOID-IC#

Consumes image classification instances without writing them.

Domain(s):#
  • Image Classification Domain
Options:#
usage: to-void-ic

TO-VOID-IS#

Consumes image segmentation instances without writing them.

Domain(s):#
  • Image Segmentation Domain
Options:#
usage: to-void-is

TO-VOID-OD#

Consumes object detection instances without writing them.

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-void-od

TO-VOID-SP#

Consumes speech instances without writing them.

Domain(s):#
  • Speech Domain
Options:#
usage: to-void-sp

TO-YOLO-OD#

Writes image object-detection annotations in the YOLO format

Domain(s):#
  • Image Object-Detection Domain
Options:#
usage: to-yolo-od [-c PATH] [-l PATH] [--annotations-only] [--no-interleave] -o PATH
                  [--split-names SPLIT NAME [SPLIT NAME ...]] [--split-ratios RATIO [RATIO ...]]

optional arguments:
  -c PATH, --labels-csv PATH
                        Path to the labels CSV file to write (default: None)
  -l PATH, --labels PATH
                        Path to the labels file to write (default: None)
  --annotations-only    skip the writing of data files, outputting only the annotation files
                        (default: False)
  --no-interleave       disables item interleaving (splitting will occur in runs) (default: False)
  -o PATH, --output PATH
                        output directory to write images and annotations to (default: None)
  --split-names SPLIT NAME [SPLIT NAME ...]
                        the names to use for the splits (default: [])
  --split-ratios RATIO [RATIO ...]
                        the ratios to use for the splits (default: [])