朱文成，天津大学人工智能学院副研究员，水木学者，硕士生导师，55教学楼B519。 2014和2017年取得天津大学学士和硕士学位，指导老师是胡清华教授和朱鹏飞教授； 2021年6月取得清华大学博士学位，导师是鲁继文长聘教授； 2021年7月清华大学水利系做博士后研究，合作导师是张建民院士。 主要关注基于视觉-语言大模型的感知、理解与强化学习策略。招收有自驱力和好奇心的硕士生，一起点亮未知，让科研成为一场充满惊喜的旅程！

I obtained my Bachelor and Master degrees from Tianjin University of Computer Science and Technology, advised by Prof. Qinghua Hu and Pengfei Zhu. I received my Doctoral degree at the Department of Automation, Tsinghua University, in Intelligent Vision Group, under the supervision of Prof. Jiwen Lu. I conducted postdoctoral research at Civil and Hydraulic Engineering， Tsinghua University, under the supervision of Prof. Jian-Min Zhang, an academician of the Chinese Academy of Engineering.

I joined the School of Artificial Intelligence at Tianjin University in November 2023, focusing on perception, reasoning, and reinforcement learning strategies of vision-language large models.

Google Scholar  /  GitHub  /  Email

Drop me an email （wenchengzhu AT tju dot edu dot cn） for the 2026 and 2027 master application !!!

Our method repurposes the frozen text encoder to construct a visual codebook from video class labels due to the many-to-one contrastive alignment between visual and textual embeddings in multimodal pretraining.

We propose a sample-wise contrastive knowledge distillation approach.

We conduct relational reasoning over spatial-temporal graphs for video summarization.

We propose a separable structure modeling approach for semi-supervised video object segmentation.

We propose a Detect-to-Summarize network for video summarization.

We propose a visual tracking approach by using set to set distance.