Domain-Adaptive Deep Joint Source-Channel Coding for Image Classification
Abstract
Deep joint source--channel coding (Deep JSCC) can suffer from domain shift between training and deployment data. We study single-source domain adaptation for image classification over wireless channels. A Gaussian example shows that improving channel quality does not necessarily improve target-domain accuracy under a fixed source-trained classifier. We further proposed an information objective to reduce domain information while preserving class information. Motivated by the analysis, we develop a domain-adaptive Deep JSCC method that combines class-wise adversarial alignment with pseudo-label supervised contrastive learning. The former uses all target samples to reduce class-conditional domain differences, while the latter selects high-confidence target samples to obtain class-discriminative received representations. We relate these losses to the information objective, and analyze the effects of pseudo-label errors and confidence-based selection. Experiments on SVHN--MNIST and PACS over AWGN and Rayleigh channels demonstrate improved target-domain classification without additional inference-time networks. The proposed method achieves 98.15\% accuracy on SVHN → MNIST at 10~dB CSNR.
Related
- Joint Source-Channel Coding of Gaussian Sources over Block Erasure Channels: Nonasymptotic Bounds and Channel-Uniform Normal Approximations
- Secure Joint Source-Channel Coding of Multimodal Semantic Sources
- Polar Complexity: A New Descriptive Complexity with Applications to Source and Joint Source-Channel Coding
- Lossy Joint Source-Channel Coding over Unknown Channels
- A Sum-of-Squares Hierarchy with Quadratic Convergence for Quantum Channel Coding
BibTeX
@misc{li2026domainadaptive,
title = {{Domain-Adaptive Deep Joint Source-Channel Coding for Image Classification}},
author = {Yishen Li and Xuechen Chen and Xiaoheng Deng and Hao Zhang},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2607.28907},
eprint = {2607.28907},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2607.28907},
}
RIS
TY - UNPB
TI - Domain-Adaptive Deep Joint Source-Channel Coding for Image Classification
AU - Yishen Li
AU - Xuechen Chen
AU - Xiaoheng Deng
AU - Hao Zhang
PY - 2026
DA - 2026/11/03
PB - arXiv
DO - 10.48550/arxiv.2607.28907
UR - https://arxiv.org/abs/2607.28907
AB - Deep joint source--channel coding (Deep JSCC) can suffer from domain shift between training and deployment data. We study single-source domain adaptation for image classification over wireless channels. A Gaussian example shows that improving channel quality does not necessarily improve target-domain accuracy under a fixed source-trained classifier. We further proposed an information objective to reduce domain information while preserving class information. Motivated by the analysis, we develop a domain-adaptive Deep JSCC method that combines class-wise adversarial alignment with pseudo-label supervised contrastive learning. The former uses all target samples to reduce class-conditional domain differences, while the latter selects high-confidence target samples to obtain class-discriminative received representations. We relate these losses to the information objective, and analyze the effects of pseudo-label errors and confidence-based selection. Experiments on SVHN--MNIST and PACS over AWGN and Rayleigh channels demonstrate improved target-domain classification without additional inference-time networks. The proposed method achieves 98.15\% accuracy on SVHN → MNIST at 10~dB CSNR.
ER -