Bio
He is a postdoc at NYU systems group working with Prof. Jinyang Li. Before that, he obtained a computer science Ph.D. at New York University where he worked with Prof. Sam Bowman. His recent research focuses on scalable continuous learning systems and efficient edge-AI systems. Prior to joining NYU, he earned his M.Tech from Purdue University and B.Tech from Sichuan University. He collaborates closely with Prof. Yuanchao Shu from Zhejiang University, Dr. Shiqi Jiang from Microsoft Research, and Prof. Huanle Zhang from Shandong University. He is the co-founder of MengJiTech (孟及科技) which is a startup focusing on AIGC.
News
- 2023.01: MVDAOD was selected as AAAI 2023 oral (10.71%)!
- 2022.11: MVDAOD was accepted to AAAI 2023! Thanks for all collaborators!!!
- 2022.11: Starting internship at Microsoft Research Shanghai!
- 2022.11: e2e-detection is released!
- 2022.10: Turbo was accepted to Sensys 2022!
- 2022.09: One paper was conditionally accepted to Sensys 2022!
- 2022.07: One paper got accepcted to TIOT 2022!
Selected Research Papers
- [AAAI 2023 (Oral)] Yan Lu, Zhun Zhong, Yuanchao Shu. “Multi-View Domain Adaptive Object Detection in Surveillance Cameras”. [arxiv] [PDF] [PDF (including appendix)] [Slides] [bib]
- Leveraging multi-views from overlapping cameras to learn a robust backbone for downstream unsupervised domain adaptation tasks.
- [SenSys 2022] Yan Lu, Shiqi Jiang, Ting Cao, Yuanchao Shu. “Turbo: Opportunistic Enhancement for Edge Video Analytics”. ACM SenSys 2022. [PDF] [Slides] [Code] [Project page] [bib]
- Using data enhancement to help video analytics applications capture hard samples and improve the accuracy in further.
- [TIOT 2022] Zheng Dong, Yan Lu, Guangmo Tong, Yuanchao Shu, Shuai Wang, Weisong Shi. “WatchDog: Real-time Vehicle Tracking on Geo-distributed Edge Nodes”. ACM Transactions on Internet of Things, 2022. [PDF] [bib]
- Accelerating vehicle tracking with a distributed resource manager!
- [SEC 2019] Yan Lu, Yuanchao Shu, Xu Tan, Yunxin Liu, Mengyu Zhou, Qi Chen and Dan Pei. “Collaborative Learning between Cloud and End Devices: An Empirical Study on Location Prediction”. In the Fourth ACM/IEEE Symposium on Edge Computing, 2019. [PDF] [bib]
- Improving edge device learning with the global knowledge!
Industry Experience
- Research Intern, Wireless Group @ Microsoft Research in Asia
- Nov 2022 - May 2023, Mentored by Dr. Shiqi Jiang
- Research Intern, Recommendation Systems Group @ Airbnb
- May 2020 - Aug 2020, Mentored by Dr. Wei Zhang
- Research Intern, Mobile and Networking Group @ Microsoft Research in Asia
- Dec 2017 - Aug 2018, Mentored by Dr. Yuanchao Shu and Dr. Yunxin Liu
- Research Assistant, Intelligent Agents Group @ Chinese Academy of Sciences
- Feb 2016 - Aug 2016, Mentored by Dr. Xinhua Zeng
Tech Blogs
- Research notes
- Base Models (language models, image classification, object detection, time-series forecasting, image synthesis)
- Continuous learning Systems (efficient training, on-device learning, continuous learning)
- Collaborative learning (device-device, cloud-device, multi-task collaboration)
- Inference Engines (edge-based and cloud-based accelerations)
- 🔥 LLM based development
- MLSys Development (frameworks, engineering tips)