Xiaochen Ma

Xiaochen Ma

Ph.D Student

MBZUAI, United Arab Emirates

Biography

Xiaochen Ma (马晓晨) is a Ph.D student in Computer vision at MBZUAI, UAE, under the mentorship of Prof. Zhiqiang Shen. Previously, he served as a full-time Research Assistant at Sichuan University, China, from July 2023 to June 2024, guided by Prof. Jizhe Zhou. Xiaochen earned his B.E. in Computer Science from the College of Computer Science at Sichuan University in June 2023, where he was awarded for his outstanding undergraduate thesis, ranking in the top 2 out of over 400 students.

He enjoys embracing the open-source community and has currently garnered 300+ stars in all of his repositories on GitHub.

His current research interest lies in Computer Vision, Image Segmentation, and Image Forensics, especially image manipulation localization. He is looking for an opportunity to develop a fundamental paradigm for image segmentation with large models.

Services:

Reviewer: ICLR'25, NeurIPS'24, and ACM MM'24, AISTATS'25.

Interests
  • Computer Vision
  • Image Forensic
  • Large Vision Model
Education
  • Ph.D in Computer Vision, 2028 (expected)

    MBZUAI

  • B.Eng. in Computer Science, 2023

    Sichuan University

Experience

 
 
 
 
 
Mohamed bin Zayed University of Artificial Intelligence
PhD Student in Computer Vision
August 2024 – Present United Arab Emirates
 
 
 
 
 
Sichuan University
Full-time Research Assistant
July 2023 – June 2024 China
 
 
 
 
 
Sichuan University
Undergraduate Student
September 2019 – June 2023 China

Publications

(2024). IMDL-BenCo: A Comprehensive Benchmark and Codebase for Image Manipulation Detection & Localization. NeurIPS'24 Spotlight.

PDF Cite Code arXiv

(2023). IML-ViT: Benchmarking Image Manipulation Localization by Vision Transformer. In Submission.

PDF Cite Code arXiv

(2023). Perceptual MAE for Image Manipulation Localization: A High-level Vision Learner Focusing on Low-level Features. In Submission.

PDF Cite Code arXiv

(2023). Pre-training-free Image Manipulation Localization through Non-Mutually Exclusive Contrastive Learning. ICCV'23.

PDF Cite Code arXiv URL

(2023). Enzymatic colorimetric method for turn-on determination of l-lactic acid through indicator displacement assay. Journal of Bioscience and Bioengineering.

Cite DOI URL

Contact

  • xiaochen.ma.cs[at]gmail.com or Xiaochen.Ma[at]mbzuai.ac.ae
  • Masdar City, Abu Dhabi, PO.Box 7909
  • Building 1B, Mohamed Bin Zayed University of Artificial Intelligence