sclee0724/PADC-DeepJSCCV-OraNet-Reproduction

PADC: DeepJSCC-V and OraNet Reproduction

Unofficial PyTorch and Google Colab reproduction of "Predictive and Adaptive Deep Coding for Wireless Image Transmission in Semantic Communication" (IEEE TWC, 2023). https://github.com/sclee0724/PADC-DeepJSCCV-OraNet

Models

• DeepJSCC-V trained on CIFAR10 (AWGN / Fading)
• OraNet trained on CIFAR10 (AWGN / Fading)
• DeepJSCC-V fine-tuned on ImageNet
• OraNet trained on ImageNet

Evaluation Datasets

• CIFAR100
• Kodak24

Checkpoints

Each DeepJSCC-V checkpoint should be paired with the OraNet checkpoint trained for the same channel and dataset configuration.

Original Implementation

https://github.com/wyzhang-ustb/Predictive-and-Adaptive-Deep-Coding-for-Wireless-Image-Transmission-in-Semantic-Communication

Citation

W. Zhang et al., "Predictive and Adaptive Deep Coding for Wireless Image Transmission in Semantic Communication," IEEE Transactions on Wireless Communications, vol. 22, no. 8, 2023.

DOI: 10.1109/TWC.2023.3234408