compressai.entropy_models

EntropyBottleneck

class compressai.entropy_models.EntropyBottleneck(channels: int, *args: Any, tail_mass: float = 1e-09, init_scale: float = 10, filters: Tuple[int, ...] = (3, 3, 3, 3), **kwargs: Any)

Entropy bottleneck layer, introduced by J. Ballé, D. Minnen, S. Singh, S. J. Hwang, N. Johnston, in “Variational image compression with a scale hyperprior”.

This is a re-implementation of the entropy bottleneck layer in tensorflow/compression. See the original paper and the tensorflow documentation for an introduction.

GaussianConditional

class compressai.entropy_models.GaussianConditional(scale_table: List | Tuple | None, *args: Any, scale_bound: float = 0.11, tail_mass: float = 1e-09, **kwargs: Any)

Gaussian conditional layer, introduced by J. Ballé, D. Minnen, S. Singh, S. J. Hwang, N. Johnston, in “Variational image compression with a scale hyperprior”.

This is a re-implementation of the Gaussian conditional layer in tensorflow/compression. See the tensorflow documentation for more information.

EntropyBottleneckVbr

class compressai.entropy_models.EntropyBottleneckVbr(channels: int, *args: Any, tail_mass: float = 1e-09, init_scale: float = 10, filters: Tuple[int, ...] = (3, 3, 3, 3), **kwargs: Any)

Entropy bottleneck layer, introduced by J. Ballé, D. Minnen, S. Singh, S. J. Hwang, N. Johnston, in “Variational image compression with a scale hyperprior”.

This is a re-implementation of the entropy bottleneck layer in tensorflow/compression. See the original paper and the tensorflow documentation for an introduction.