|
I am currently a Ph.D. student in School of Software, Tsinghua University, advised by Prof. Junhai Yong and Bin Wang. I am expected to graduate in 2027. My research interests center on computer vision and generative models, with a particular focus on diffusion models. Within this area, I work on efficient inference (TAP, DiffSparse, DiP-GO, TSS) and data augmentation (ReCon, DistDiff). Additionally, I explore parameter-efficient fine-tuning (SHIP) and fine-grained visual recognition (DCAL). |
|
|
|
|
 |
Haowei Zhu, Tingxuan Huang, Xing Wang, et al. CVPR, 2026 arXiv Training-free framework for adaptive token-level prediction in diffusion acceleration. |
 |
Haowei Zhu, Ji Liu, Ziqiong Liu, et al. ICLR, 2026 arXiv A differentiable sparsity framework to accelerate few-step diffusion transformers. |
 |
Haowei Zhu, Tianxiang Pan, Rui Qin, et al. NeurIPS, 2025 (Spotlight) arXiv | code Region-controllable data augmentation with rectification and alignment for object detection. |
 |
Haowei Zhu, Dehua Tang, Ji Liu, et al. NeurIPS, 2024 arXiv Differentiable pruning network with few-step gradient optimization for efficient diffusion model acceleration. |
 |
Haowei Zhu, Ling Yang, Junhai Yong, et al. NeurIPS, 2024 arXiv | code Hierarchical prototypes and distribution-aware energy guidance for diffusion sampling. |
 |
Haowei Zhu, Wenjing Ke, Dong Li, et al. CVPR, 2022 arXiv Global-local cross-attention and pairwise attention modules for fine-grained tasks. |
|