kaira.models.image.DeepJSCCFeedbackDecoder

Inheritance diagram of DeepJSCCFeedbackDecoder

Inheritance diagram for DeepJSCCFeedbackDecoder

class kaira.models.image.DeepJSCCFeedbackDecoder(n_channels: int, *args: Any, **kwargs: Any)[source]

Bases: BaseModel

Decoder network for DeepJSCC with Feedback [Kurka and Gündüz, 2020].

This decoder reconstructs the image from the received noisy channel output. The architecture uses transposed convolutions with inverse GDN activations to convert the channel signal back into an image.

Parameters:

n_channels (int) – Number of channels in the output image (typically 3 for RGB).

Methods

__init__

Initialize the DeepJSCCFeedbackDecoder.

forward

Forward pass through the decoder.

from_config

Create model instance from configuration.

from_hydra_config

Create model from Hydra DictConfig.

from_pretrained_config

Create model from Hugging Face PretrainedConfig.
__init__(n_channels: int, *args: Any, **kwargs: Any)[source]

Initialize the DeepJSCCFeedbackDecoder.

Parameters:

  • n_channels (int) – Number of channels in the output image.

  • *args – Variable positional arguments passed to the base class.

  • **kwargs – Variable keyword arguments passed to the base class.

forward(x: Tensor, *args: Any, **kwargs: Any) → Tensor[source]

Forward pass through the decoder.

Parameters:

  • x (torch.Tensor) – Channel output tensor to be decoded.

  • *args – Additional positional arguments (passed to internal layers).

  • **kwargs – Additional keyword arguments (passed to internal layers).

Returns:

Reconstructed image in range [0, 1].

Return type:

torch.Tensor

classmethod from_config(config, **kwargs)

Create model instance from configuration.

Parameters:

  • config – Configuration object (PretrainedConfig, DictConfig, or dict)

  • **kwargs – Additional parameters to override config

Returns:

Model instance

classmethod from_hydra_config(config: DictConfig, **kwargs)

Create model from Hydra DictConfig.

Parameters:

  • config – Hydra configuration

  • **kwargs – Additional parameters

Returns:

Model instance

classmethod from_pretrained_config(config: PretrainedConfig, **kwargs)

Create model from Hugging Face PretrainedConfig.

Parameters:

  • config – PretrainedConfig instance

  • **kwargs – Additional parameters

Returns:

Model instance