Welcome to my Homepage!
Hi! I am a second-year master’s student and I am currently studying at National University of Defense Technology with Prof. Zhiping Cai. My current research focuses on Style Transfer, 3D Reconstruction Based on 3DGS and Edge Detection. Feel free to reach out if you are interested in collaboration or potential opportunities.
News
- 2026.09 🎉🎉 One paper has successfully advanced to the second stage of AAAI27.
- 2026.07 🎉🎉 One paper has been submitted to AAAI 2027.
- 2026.06 🎉🎉 I began my internship at Huawei.
- 2026.03 🎉🎉 StyleGallery has been accepted by CVPR 2026.
- 2026.02 🤡🤡 IJCAI has been downgraded to CCF-B
- 2026.02 🎉🎉 SCGS has been submitted to IJCAI 2026.
- 2025.11 🎉🎉 StyleGallery has been submitted to CVPR 2026.
- 2024.09 🎉🎉 I began my graduate studies at NUDT.
- 2024.06 🎉🎉 I have successfully completed my undergraduate studies!.
Experience
Huawei Nanjing Research Institute
2026.6 - 2026.9 (Outstanding Intern)
Haisi Semiconductor Business Department, Internship
2026.6 - 2026.9 (Outstanding Intern)
Haisi Semiconductor Business Department, Internship
National University of Defense Technology
2024.09 - Present
College of Computer Science, Master's Degree Candidate
2024.09 - Present
College of Computer Science, Master's Degree Candidate
Ocean University of China
2020.09 - 2024.06
Department of Information Science and Engineering, Bachelor's Degree
2020.09 - 2024.06
Department of Information Science and Engineering, Bachelor's Degree
Publications
(* equal contribution · † corresponding author · ‡ project leader)
StyleGallery: Training-free and Semantic-aware Personalized Style Transfer from Arbitrary Image References
Boyu He* , Yunfan Ye* , Chang Liu, Weishang Wu, Fang Liu†, Zhiping Cai†.
We introduce StyleGallery, a training-free and semantic-aware framework for personalized style transfer from arbitrary image references. By adaptively clustering and matching semantic regions without extra masks, it enables fine-grained, interpretable stylization while preserving global content structure and supporting flexible customization with multiple style references.
CVPR 2026 [arXiv] [code]
Boyu He* , Yunfan Ye* , Chang Liu, Weishang Wu, Fang Liu†, Zhiping Cai†.
We introduce StyleGallery, a training-free and semantic-aware framework for personalized style transfer from arbitrary image references. By adaptively clustering and matching semantic regions without extra masks, it enables fine-grained, interpretable stylization while preserving global content structure and supporting flexible customization with multiple style references.
CVPR 2026 [arXiv] [code]