I'm Ma Xinran.
Sichuan University, China
Welcome to XLearning Lab!
My name is Xinran Ma. Now I am studying for my master’s degree in College of Computer Science, Sichuan University. I work in XLearning Laboratory in Basic Teaching Building B316, Wangjiang Campus, Sichuan University. My tutor is Professor Peng Xi. I am very glad to be a member of the scientific research staff, and welcome you to visit me and our laboratory.
Research Interests
Multi-Modal Learning, Learning with Noisy Correspondence
News
[2024-02] Feb. 28, 2024, one paper was accepted by IEEE Transactions on Image Processing (TIP)!
[2022-06] I graduated from the Wu Yuzhang Honors College of Sichuan University in June 2022 with a Bachelor of Engineering degree in Computer Science and Technology.
Publications
[IEEE TIP 2024] Xinran Ma#, Mouxing Yang#, Yunfan Li, Peng Hu, Jiancheng Lv, Xi Peng*, Cross-modal Retrieval with Noisy Correspondence via Consistency Refining and Mining, IEEE Transactions on Image Processing (TIP), 2024, DOI: 10.1109/TIP.2024.3374221
[Abstract] The success of existing cross-modal retrieval (CMR) methods heavily rely on the assumption that the annotated cross-modal correspondence is faultless. In practice, however, the correspondence of some pairs would be inevitably contaminated during data collection or annotation, thus leading to the so-called Noisy Correspondence (NC) problem. To alleviate the influence of NC, we propose a novel method termed Consistency REfining And Mining (CREAM) by revealing and exploiting the difference between correspondence and consistency. Specifically, the correspondence and the consistency only be coincident for true positive and true negative pairs, while being distinct for false positive and false negative pairs. Based on the observation, CREAM employs a collaborative learning paradigm to detect and rectify the correspondence of positives, and a negative mining approach to explore and utilize the consistency. Thanks to the consistency refining and mining strategy of CREAM, the overfitting on the false positives could be prevented and the consistency rooted in the false negatives could be exploited, thus leading to a robust CMR method. Extensive experiments verify the effectiveness of our method on three image-text benchmarks including Flickr30K, MS-COCO, and Conceptual Captions. Furthermore, we adopt our method into the graph matching task and the results demonstrate the robustness of our method against fine-grained NC problem.
Activities
[2021] Outstanding Graduate, Sichuan University
[2021,2020,2019] Outstanding Student, Sichuan University
[2021,2020,2019] Comprehensive Scholarship, Sichuan University