IDEAL: In-DEpth ALignment Makes A Discrete Representation AutoEncoder
Researchers have introduced IDEAL, an In-depth Alignment framework designed to improve discrete representation autoencoders (RAEs) for image generation. By combining both shallow and deep features from vision foundation models (VFMs), IDEAL enhances the preservation of fine-grained visual detail and semantic richness. This approach leads to superior reconstruction performance, achieving a new state-of-the-art rFID score of 0.61 on ImageNet and a gFID of 1.89 for autoregressive image generation. AI
IMPACT Enhances image generation quality by preserving both visual fidelity and semantic richness in discrete representations.