City University of Hong Kong · 5 records

dateidtitleauthorsvenue↑
2026-10-26W4416092020TokLIP: Marry Visual Tokens to CLIP for Multimodal Comprehension and GenerationHaokun Lin et al.International Journal of Computer Vision
2026-09-252609.23087Neural Spectral Capacity: Measuring and Designing Architectures from Network Specification AloneChenyu Zhu et al.arXiv11
2026-09-212609.20633Refinement Is Inherently Editable: Training-Free Prompt-to-Prompt Image Editing with Generative Refinement NetworkYulong Chen et al.arXiv9
2026-08-112608.07463MirrorWorld: Taming Video Diffusion Models for Mirror Reflection GenerationYoujun Zhao et al.arXiv3
2026-06-172606.18180EgoCS-400K: An Egocentric Gameplay Dataset for World ModelsRongjin Guo et al.arXiv17