Scholar
Yunsheng Li
Google Scholar ID: hJrIyCwAAAAJ
Microsoft
Citations & Impact
Citations
3,907
H-index
12
i10-index
13
Publications
20
Co-authors
19
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone
- Rethinking Visual Prompting for Multimodal Large Language Models with External Knowledge
- SCHEME: Scalable Channel Mixer for Vision Transformers
- Fully Authentic Visual Question Answering Dataset from Online Communities
- Dense Network Expansion for Class Incremental Learning
- Should All Proposals Be Treated Equally in Object Detection?
- MicroNet: Towards Image Recognition with Extremely Low FLOPs
- Dynamic Transfer for Multi-Source Domain Adaptation
- Revisiting Dynamic Convolution via Matrix Decomposition
- Explainable Object-Induced Action Decision for Autonomous Vehicles
- Bidirectional Learning for Domain Adaptation of Semantic Segmentation
- Efficient Multi-Domain Learning by Covariance Normalization
- Deep Scene Image Classification with the MFAFVNet
- Deep Hashing with Hash-Consistent Large Margin Proxy Embeddings
- Semantic Fisher Scores for Task Transfer: Using Objects to Classify Scenes
Education
- 2015-2021: Ph.D. student at the University of California, San Diego, focusing on overcoming resource-constrained computer vision topics such as efficient neural network architecture design and domain adaptation.
Background
- Yunsheng Li is a Senior Researcher at Microsoft Azure GenAI Group. He is working on the development of multi-modality large language models. His research interests include computer vision (segmentation, domain adaptation), deep learning (network architecture design), and multi-modality large language models. His representative works include phi-3-vision, MicroNet, and BDL.
