Wenze Liu

6

Papers

62

Total Citations

Papers (6)

DiTCtrl: Exploring Attention Control in Multi-Modal Diffusion Transformer for Tuning-Free Multi-Prompt Longer Video Generation

Unifying Automatic and Interactive Matting with Pretrained ViTs

Training Matting Models Without Alpha Labels

Learning to Upsample by Learning to Sample

FADE: Fusing the Assets of Decoder and Encoder for Task-Agnostic Upsampling

SAPA: Similarity-Aware Point Affiliation for Feature Upsampling