Jiasheng Tang

6

Papers

44

Total Citations

Papers (6)

Dynamic Diffusion Transformer

FPSAttention: Training-Aware FP8 and Sparsity Co-Design for Fast Video Diffusion

SparseDiT: Token Sparsification for Efficient Diffusion Transformer

A Stitch in Time Saves Nine: Small VLM is a Precise Guidance for Accelerating Large VLMs

ShapeBoost: Boosting Human Shape Estimation with Part-Based Parameterization and Clothing-Preserving Augmentation

Accelerating Parallel Sampling of Diffusion Models