教育立根之本
随着计算机科学与技术在人们生活中的应用不断深入,尤其是云计算、 物联网、人工智能、大数据等技术的兴起, 人工智能专业人才的需求不断提高
上海交通大学长聘教轨助理教授、博士生导师,入选国家高层次青年人才计划、上海市海外高层次青年人才计划。本科毕业于复旦大学,博士毕业于加州大学伯克利分校,师从 Trevor Darrell 教授。主要研究方向为科学智能(AI for Science)和多模态大语言模型(Multimodal Large Language Models),研究工作发表在 ICLR、ICML、NeurIPS、CVPR、ICCV、ECCV、ICRA 等国际顶级会议,近五年论文谷歌学术总引用次数 10000 余次。更多信息请访问更多信息请访问 https://dequan.wang
本科生课程《人工智能算法实践》
研究生课程《设计与理解深度神经网络》
Gao, Jin, et al. "Dissecting Dissonance: Benchmarking Large Multimodal Models Against Self-Contradictory Instructions." European Conference on Computer Vision (ECCV), 2024.
Zhao, Juntu, et al. "Lost in Translation: Latent Concept Misalignment in Text-to-Image Diffusion Models." European Conference on Computer Vision (ECCV), 2024.
Wang, Dequan, et al. "A Real-world Dataset and Benchmark For Foundation Model Adaptation in Medical Image Classification." Scientific Data. 2023.
Gao, Jin, et al. "Back to the source: Diffusion-driven adaptation to test-time corruption." IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2023.
Zhang, Yunkun, et al. "Text-guided foundation model adaptation for pathological image classification." International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI). 2023.
国家自然科学基金优秀青年科学基金项目(海外)
山东省重点研发计划(重大科技创新工程)
国家高层次青年人才
上海海外高层次青年人才
世界人工智能大会云帆奖·明日之星
Organizer of CVPR 2024 Workshop: Test-Time Adaptation Workshop
Organizer of NeurIPS 2023 Challenge: Foundation Model Prompting for Medical Image Classification Challenge
Guest Editor of MIA Special Issue: Foundation Models for Medical Image Analysis
地址:上海市闵行区东川路800号电院群楼3号楼301室
邮编:200240
电话:021 – 34204113
邮箱:qingyuan@sjtu.edu.cn