AI RESEARCH
Dual-Latent Collaborative Decoding for Fidelity-Perception Balanced Image Compression
arXiv CS.CV
•
ArXi:2605.14391v1 Announce Type: new Learned image compression (LIC) increasingly requires reconstructions that balance distortion fidelity and perceptual realism across a wide range of bitrates. However, most existing methods still rely on a single compressed latent representation to simultaneously carry structural details, semantic cues, and perceptual priors, requiring the same latent representation to serve multiple, potentially conflicting roles.