New framework enhances Deep JSCC for image classification across domains

Researchers have developed a novel domain-adaptive framework for Deep Joint Source-Channel Coding (Deep JSCC) to improve image classification performance under distribution shifts. The proposed Classification-Capacity-Invariance (CCI) function analyzes how channel capacity and cross-domain invariance impact accuracy. Experiments on digit and PACS datasets demonstrated improved generalization over AWGN and Rayleigh fading channels, achieving 98.15% accuracy on SVHN to MNIST transfer at 10 dB CSNR. AI

IMPACT This research could lead to more robust image classification systems in real-world scenarios with varying data distributions.

RANK_REASON The cluster contains a research paper detailing a new method for Deep JSCC. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• AWGN
• Classification-Capacity-Invariance (CCI)
• Deep JSCC
• MNIST
• SVHN

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →