Hengjia Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Hengjia Li has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Boxi Wu
Zhejiang University
2 shared publications - Can Ye
Ant Group, Hangzhou, China
1 shared publication - Chaotian Song
College of Software, Zhejiang University, Hangzhou, China
1 shared publication - Deng Cai
State Key Lab of CAD&CG, Zhejiang University, Hangzhou, China
1 shared publication - Di Weng
State Key Lab of CAD&CG, Zhejiang University, Hangzhou, China
1 shared publication - Fan Zhuo
College of Computer Science and Technology, Zhejiang University, Hangzhou, China
1 shared publication
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 3017-3026
Probing Human-Centric Rule-based Reasoning Abilities of Video Generation Models for Vision Foundation Intelligence
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 3017-3026
Open AccessProbing Human-Centric Rule-based Reasoning Abilities of Video Generation Models for Vision Foundation Intelligence
- Top Cited Papers
- 20 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 34 (10577149) , 8145-8158
LoRA-Composer: Leveraging Low-Rank Adaptation for Multi-Concept Customization in Training-Free Diffusion Models
Customization generation techniques have significantly advanced the synthesis of specific concepts across varied contexts. Multi-concept customization emerges as the challenging task within this domain. Existing approaches often rely on training a fusion matrix of multiple Low-Rank Adaptations (LoRAs) to merge various concepts into a single image. However, we identify this straightforward method faces two major challenges: 1) concept confusion, where the model struggles to preserve distinct individual characteristics, and 2) concept vanishing, where the model fails to generate the intended subjects. To address these issues, we introduce LoRA-Composer, a training-free framework designed for seamlessly integrating multiple LoRAs, thereby enhancing the harmony among different concepts within generated images. LoRA-Composer addresses concept vanishing through concept injection constraints, enhancing visibility via an expanded cross-attention mechanism. To combat concept confusion, concept isolation constraints are introduced, refining the self-attention computation. Furthermore, we propose two inference techniques to accelerate inference speed without performance degradation and enhance the accuracy of the generated region, respectively. Extensive experiments demonstrate that LoRA-Composer significantly outperforms standard baselines, especially in scenarios without image-based conditions such as canny edge or pose estimation.
- 20 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 34 (10577149) , 8145-8158
Top Cited PapersLoRA-Composer: Leveraging Low-Rank Adaptation for Multi-Concept Customization in Training-Free Diffusion Models
Customization generation techniques have significantly advanced the synthesis of specific concepts across varied contexts. Multi-concept customization emerges as the challenging task within this domain. Existing approaches often rely on training a fusion matrix of multiple Low-Rank Adaptations (LoRAs) to merge various concepts into a single image. However, we identify this straightforward method faces two major challenges: 1) concept confusion, where the model struggles to preserve distinct individual characteristics, and 2) concept vanishing, where the model fails to generate the intended subjects. To address these issues, we introduce LoRA-Composer, a training-free framework designed for seamlessly integrating multiple LoRAs, thereby enhancing the harmony among different concepts within generated images. LoRA-Composer addresses concept vanishing through concept injection constraints, enhancing visibility via an expanded cross-attention mechanism. To combat concept confusion, concept isolation constraints are introduced, refining the self-attention computation. Furthermore, we propose two inference techniques to accelerate inference speed without performance degradation and enhance the accuracy of the generated region, respectively. Extensive experiments demonstrate that LoRA-Composer significantly outperforms standard baselines, especially in scenarios without image-based conditions such as canny edge or pose estimation.
- Open Access
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 12227-12236
Singing Timbre Popularity Assessment Based on Multimodal Large Foundation Model
- 27 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 12227-12236
Open AccessSinging Timbre Popularity Assessment Based on Multimodal Large Foundation Model