Adaptive cross-channel transformation based on self-modulation for learned image compression

Open Webpage

Wen Tan, Youneng Bao, Fanyang Meng, Chao Li, Yongsheng Liang

Published: 01 Jan 2025, Last Modified: 15 May 2025Signal Process. Image Commun. 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Analyze cross-channel nonlinear transformation to overcome GDN’s limitations.•Use self-modulation to enable adaptive and fine-grained nonlinear representation.•Propose a cross-channel transformation using spatial-channel correlation for LIC.•Improve rate-distortion performance by enhancing nonlinear representation capacity.