Welcome to the Group of Multimedia (MM) @ College of Intelligence and Computing, Tianjin University, China. Our research has been mainly in the basic theories, efficient algorithms and application of artificial intelligence and machine learning. We mainly focus on cross-media perception, understanding and reasoning, as well as robust intelligent learning models. Aiming at the semantic gap in multimedia data, we explore visual and semantic understanding of image, video, and text data. Our research interests also include the adversarial attack and robustness improvement of deep neural networks on data containing natural noises or adversarial noises to cope with low generalization ability and high vulnerability of machine learning methods under noisy scenarios. Recently, we are revolving around intelligent learning and games for open domains, including machine learning aided by “cloud-terminal” cooperation, privacy-preserving domain adaptation and adversarial games. The central goal of our researches is to improve the interpretability and robustness of deep learning models based on the theoretical basis of cross-media understanding and adversarial machine learning, so as to build more reliable multimedia and intelligent learning systems.
We have published papers on leading journals and conferences of multimedia, computer vision, machine learning, and artificial intelligence, such as IEEE TPAMI, IEEE TIP, IEEE TKDE, IEEE TNNLS, IEEE TMM, IEEE TCSVT, IEEE TCYB, ACM MM, CVPR, ICCV, ECCV, NeurIPS, AAAI, IJCAI, etc. Our paper was awarded “Best Paper Finalist” of ACM Multimedia 2017. We also got winner records in main technical challenges such as the Champion of the Large Scale Movie Description Challenge (LSMDC 2017, joint with ICCV 2017) and the Runner-up of the 2nd MSR Large-Scale Video to Language Challenge (Honorable Mention Award of Grand Challenge @ ACM MM 2017).
We are looking for passionate new PhD students and Master students to join the team !
If you are interested, please contact Prof. Yahong Han
Undergraduate courses:
Media Computing
Mingshi's paper "Behavior-Contextualized Item Preference Network for Multi-Behavior Recommendation" was accepted by SIGIR 2024 (CCF-A).
06. March 2024Runhua's paper "Generalizing to Out-of-Sample Degradations via Model Reprogramming" was accepted by IEEE TIP (CCF-A).
29. February 2024Yucheng's paper "The Metric System and Its Completeness of Adversarial Robustness Evaluation" was accepted by Journal of Software (软件学报).
27. February 2024Deng Li's paper "Prompt-Driven Dynamic Object-Centric Learning for Single Domain Generalization" was accepted by CVPR 2024 (CCF-A).
29. December 2023Yidan Fan's paper "Weakly-Supervised Video Anomaly Detection with Snippet Anomalous Attention" was accepted by IEEE TCSVT.
26. December 2023Nana Yu's paper "Joint Correcting and Refinement for Balanced Low-Light Image Enhancement" was accepted by IEEE TMM.
17. December 2023Boyuan Zhang's paper "Cascade & Allocate: A Cross-structure Adversarial Attack against Models Fusing Vision and Language" was accepted by Information Fusion.