July 15, 2025

2025 MSRA StarTrack Seminar

Location: Meeting Room 74, Building 1, Beijing office, Microsoft Research Asia

Time	Session	Speaker
13:30 – 13:40	Opening and Welcome	Baining Guo, Microsoft Research Asia
13:40 – 14:10	Invited MSRA Research Talks (2×15 mins)	Jiaolong Yang, Microsoft Research Asia Li Zhang, Microsoft Research Asia
14:10 – 15:40	Idea Sparks Panel I (From Vision to Deployment: Scaling Multimodal Foundation Models for Real-World Impact)	Chong Luo, Microsoft Research Asia (Host)
		Chong Luo, Microsoft Research Asia (Host)
	The Research Journey: From Multimedia to Multimodal, then Multi-what?	Di Hu, Renmin University of China Collaborative researcher: Jianlong Fu, Microsoft Research Asia
	Towards reasoning multimodal LLMs in low-resource scenarioes	Guanhua Chen, Southern University of Science and Technology Collaborative researcher: Dongdong Zhang, Microsoft Research Asia
	Towards Generalizable Human-Level Multimodal Generalist	Hao Fei, National University of Singapore Collaborative researcher: Lei Cui, Microsoft Research Asia
	When Do Multimodal Foundation Models Need 3D Capabilities	Pengshuai Wang, Peking University Collaborative researcher: Jiaolong Yang, Microsoft Research Asia
	Toward Self-Supervised Large Feedforward Systems	Tong Zhang, University of Chinese Academy of Sciences Collaborative researcher: Baining Guo, Microsoft Research Asia
	Fostering Digital Trust: Combating Untrustworthy Information with Multimodal AI	Yupeng Li, Hong Kong Baptist University Collaborative researcher: Fangzhao Wu, Microsoft Research Asia
	Information foraging with multimodal LLMs: opportunities and challenges	Ziang Xiao, Johns Hopkins University Collaborative researcher: Xiaoyuan Yi, Microsoft Research Asia
	Beyond Behavioral Alignment: Toward Neural-Level Alignment in Multimodal Foundation Models	Ziyu Jia, Institute of Automation of Chinese Academy of Sciences Collaborative researcher: Yansen Wang, Microsoft Research Asia
	Discussion	All
15:40 – 16:00	Group Photo and Tea Break	All
16:00 – 17:30	Idea Sparks Panel II (Understanding efficiency through the lens of intelligence)	Fan Yang, Microsoft Research Asia (Host)
	Observations on the Evolving of LLM Intelligence and Efficiency	Yuqing Yang, Microsoft Research Asia
	Balancing Efficiency and Intelligence in Speech Enhancement: Insights from Recent Advances	Chenda Li, Shanghai Jiao Tong University Collaborative researcher:Shujie Liu, Microsoft Research Asia
	Designing system Software for Wafer-Scale AI computing	Luo Mai, University of Edinburgh (Online) Collaborative researcher: Fan Yang, Microsoft Research Asia
	Flexible Sensing for Physical Perception: Gaining Efficiency in Embodied AI	Minhui Xie, Renmin University of China Collaborative researcher: Ran Shu, Microsoft Research Asia Baotong Lu, Microsoft Research Asia
	Tackling Data Redundancy in the Generative AI Era	Yihao Chen, Tsinghua University Collaborative researcher: Zilong Wang, Microsoft Research Asia Lili Qiu, Microsoft Research Asia
	Rethinking the efficiency of generative models	Zhenghao Chen, The University of Newcastle, Australia (Online) Collaborative researcher: Bin Li, Microsoft Research Asia
	Discussion	All
17:30 – 17:40	Closing Remarks	Lily Sun, Microsoft Research Asia
18:00 – 20:00	Dinner + Question Box Interaction