Nan Duan

Tech Fellow at StepFun

Dr. Nan Duan, Tech Fellow at StepFun, leads a team of researchers to build multimodal fundamental models centered on language and video. Previously, he was a Senior Principal Researcher and Research Manager of the Natural Language Computing team at Microsoft Research Asia (2012-2024). Dr. Duan is an Adjunct Doctoral Professor at the University of Science and Technology of China and Xi'an Jiaotong University, and an Adjunct Professor at Tianjin University. He focuses on natural language processing, code intelligence, multimodal base models, and intelligences.

Topic

Advances, Challenges, and the Future of Video Generation Foundation Modeling

This report will present the latest advances in basic video generation models, including text-generated video and graph-generated video tasks, centered on the Step-Video family of open-source models. In addition, it will summarize the main challenges faced by existing video generation models and discuss possible future directions.

© boolan.com 博览 版权所有

沪ICP备15014563号-6

沪公网安备31011502003949号