About
A computer vision researcher interested in both long-term research and short-term biz-landing.
Now: I am focused on human-centric image and video generation technologies at Meta BizAI. Meanwhile, I am building foundation models capable of understanding and generating across diverse modalities.
Previously: I received my Ph.D. in Computer Science from Leibniz Hanover University, advised by Prof. Michael Yang and Prof. Bodo Rosenhahn. I was a senior ML scientist at Picsart, a research intern at Meta and Bosch. I obtained my Master's degree from Leibniz Hanover University and my Bachelor's degree from Hefei University of Technology.
News
- Our GenAI solution for Ads is highlighted at Cannes Lions 2025.
- Paper on compositional image generation accepted to IJCV.
- Paper on virtual try-on accepted to CVPR.
Selected Publications
Generative Models
- 
               Attribute-Centric Compositional Text-to-Image Generation Attribute-Centric Compositional Text-to-Image Generation
- 
               Learning Flow Fields in Attention for Controllable Person Image Generation Learning Flow Fields in Attention for Controllable Person Image Generation
- 
                
- 
                
Scene Understanding
- 
               SPAN: Learning Similarity between Scene Graphs and Images with Transformers SPAN: Learning Similarity between Scene Graphs and Images with Transformers
- 
                
- 
                
- 
                
Embodied AI
- 
               Indoor Scene Change Understanding (SCU): Segment, Describe, and Revert Any Change Indoor Scene Change Understanding (SCU): Segment, Describe, and Revert Any Change
- 
               Worldafford: Affordance Grounding Based on Natural Language Instructions Worldafford: Affordance Grounding Based on Natural Language Instructions
Contact
- Email: congyuren@hotmail.com
- Wechat: congyr5