Semantic image communication improves efficiency using region graphs
Channel-Adaptive Region Adjacency Graph Carriers for Semantic Image Communication
Machine Learning
Summary
Sending images over noisy channels can be tricky because important parts might get lost or distorted. The authors developed a method that breaks images into meaningful regions connected in a graph, which helps keep important scene details even when data is limited or noisy. Their approach makes the data needed smaller and keeps the key details better than some other methods. This can help in tasks where understanding image content matters more than just looking nice.
What this means in practice
- •For embedded system developers: Design image transmission systems that maintain important scene details with fewer channel resources using semantic region graphs.
- •For wireless communication engineers: Improve robustness of image data sent over noisy wireless channels by encoding adjacency and region information instead of full images.
Authors
Karim Abdallah, Maria Slim, Mariette Awad, Hadi Sarieddeen
Abstract
Semantic image communication seeks to preserve task-relevant scene structure under limited channel resources, but carriers are often dense latent tensors or grid-aligned semantic layouts that do not explicitly encode region-level relations. This work introduces a segmentation-derived region adjacency graph (RAG) carrier, termed channel-adaptive RAG (CA-RAG), for joint source-channel coding-style image communication. Nodes store interpretable region attributes, edges preserve adjacency, channel-adaptive graph simplification (CGS) controls the node budget, and semantic belief propagation refines noisy graph embeddings before diffusion-based reconstruction. On Cityscapes, pre-channel RAG payloads are several times smaller than compressed class-index layouts in a 2,000-image study. Under additive white Gaussian noise at signal-to-noise ratios from 0 to 15 dB, CA-RAG reports higher semantic consistency than deep joint source-channel coding and a same-decoder layout baseline, with comparable perceptual quality. At 10 dB, the full-budget rate-sweep point reaches mean intersection over union (mIoU) = 0.329 at approximately 3.3 x 10^3 channel uses, while the default adaptive-CGS setting reports mIoU = 0.294 at approximately 2.6 x 10^3 channel uses.