Example¶

Encoding an image into a .cool bitstream¶

Encoding an image with CoolChic requires to specify

  • The input image path -i and output bitstream path -o

  • A working directory --workdir where logs are written.

  • The encoder configuration --enc_cfg for the training options.

  • The decoder configuration --dec_cfg for the neural networks architecture.

(venv) ~/Cool-Chic$ python coolchic/encode.py \
    -i=image.png -o=./bitstream.cool --workdir=./dummy_workdir \
    --enc_cfg=cfg/enc/fast.cfg --dec_cfg=cfg/dec/lop.cfg

More details on encoding images with Cool-chic is available in the encoder documentation.

Decoding a .cool bitstream¶

Decoding an image with CoolChic requires to specify

  • The input bitstream path -i and the decoded image path -o

(venv) ~/Cool-Chic$ python coolchic/decode.py -i=bitstream.cool -o=decoded_image.ppm

Note that Cool-Chic outputs either PPM or YUV files.

More details on decoding images with Cool-chic is available in the decoder documentation.