11 papers
LESA: Learnable Stage-Aware Predictors for Diffusion Model Acceleration
Peiliang Cai, Jiacheng Liu, Haowen Xu +3
Diffusion models have achieved remarkable success in image and video generation tasks. However, the high computational demands of Diffusion Transformers (DiTs) pose a significant c…
Dynamic Video Generation: Shaping Video Generation Across Time and Space
Shikang Zheng, Jingkai Huang, Jiacheng Liu +5
Diffusion models have achieved impressive performance in video generation, but their iterative denoising process remains computationally expensive due to the large number of tokens…
Focused Forcing: Content-Aware Per-Frame KV Selection for Efficient Autoregressive Video Diffusion
Peiliang Cai, Evelyn Zhang, Jiacheng Liu +8
Recent advances in autoregressive video diffusion have enabled sequential and streaming video generation. However, long-horizon generation requires increasingly large KV caches, ma…
Bridging Generation and Training: A Systematic Review of Quality Issues in LLMs for Code
Kaifeng He, Xiaojun Zhang, Peiliang Cai +7
Large language models (LLMs) frequently generate defective outputs in code generation tasks, ranging from logical bugs to security vulnerabilities. While these generation failures…
SpecEdit: Training-Free Acceleration for Diffusion based Image Editing via Semantic Locking
Zhengan Yan, Shikang Zheng, Haoran Qin +9
Diffusion-based image editing offers strong semantic controllability, but remains computationally expensive due to iterative high-resolution denoising over all spatial tokens. Dyna…
Beyond Fixed Formulas: Data-Driven Linear Predictor for Efficient Diffusion Models
Zhirong Shen, Rui Huang, Jiacheng Liu +6
To address the high sampling cost of Diffusion Transformers (DiTs), feature caching offers a training-free acceleration method. However, existing methods rely on hand-crafted forec…