Submitted by akhaliq 25 CatLIP: CLIP-level Visual Recognition Accuracy with 2.7x Faster Pre-training on Web-scale Image-Text Data · 8 authors 3
Submitted by akhaliq 16 PuLID: Pure and Lightning ID Customization via Contrastive Alignment · 5 authors 1
Submitted by akhaliq 11 ID-Aligner: Enhancing Identity-Preserving Text-to-Image Generation with Reward Feedback Learning · 6 authors 1
Submitted by akhaliq 10 MotionMaster: Training-free Camera Motion Transfer For Video Generation · 8 authors 1
Submitted by akhaliq 7 XC-Cache: Cross-Attending to Cached Context for Efficient LLM Inference · 8 authors 1