Foveated image compression from a semantic communication perspective

This project is for foveated compression within a semantic framework. It leverages the biological principle of foveation where high visual quality is concentrated at the gaze center by integrating it with semantic compression. This dual approach will encode critical data in the user's central vision while aggressively compressing the peripheral visual field. The end goal will be dramatic reduction in required bit rates. This project will be primarily targeted for AR/VR applications, enabling higher-quality, lower-latency, and more bandwidth-efficient immersive experiences, which is crucial for widespread adoption and advanced functionality.