Wenxin (Wendy) Ma 马雯芯
I’m Wendy, a first-year PhD student in Computer Science at The University of Texas at Austin, advised by Prof. Philipp Krähenbühl. My research spans computer vision and medical image analysis, with an interest in active exploration and spatial and temporal reasoning.
I joined the MIRACLE Lab at the University of Science and Technology of China (USTC) for my master’s studies in 2023, advised by Prof. S. Kevin Zhou and Prof. Zihang Jiang. In April 2025, I joined Johns Hopkins University as a visiting scholar, working with Prof. Alan Yuille and Jieneng Chen.
Previously, I received my bachelor’s degree from the Zhiyuan Honored Program at Shanghai Jiao Tong University.
Research interests
I am interested in how AI agents actively explore their environments, gather useful information, and reason about how the world changes over time. My research focuses on:
- Active exploration: enabling agents to choose actions and observations that reduce uncertainty and build a better understanding of their environments.
- Spatial and temporal reasoning: learning representations that capture 3D structure, remain consistent over time, and support plausible cause-and-effect reasoning.
- Medical image analysis and anomaly detection: developing representations that help identify unusual patterns and support image segmentation.
Selected publications
† Equal contribution · * Corresponding author. Full publication list on Google Scholar ↗
Leading works
First-author and co-first-author research.
2026
Spatial Code
Help language models reason about the physical world by turning video into explicit 3D representations.
Thinking with Spatial Code for Physical-World Video Reasoning
Wenxin MaCo-first author
CausalSpatial
Test whether multimodal models can predict what happens when an object moves—across collision, compatibility, occlusion, and trajectory tasks.
CausalSpatial: A Benchmark for Object-Centric Causal Spatial Reasoning
Wenxin MaFirst author
2025
AA-CLIP
Make CLIP representations more sensitive to anomalies for zero-shot detection.
AA-CLIP: Enhancing Zero-shot Anomaly Detection via Anomaly-Aware CLIP
Wenxin MaFirst author
LLM4Seg
Use a frozen pretrained language-model layer to strengthen visual representations for medical image segmentation.
Pre-Trained LLM is a Semantic-Aware and Generalizable Segmentation Booster
Wenxin MaCo-first author
UniAS
Bring anomaly segmentation into a unified approach that identifies unusual regions in images.
Towards Accurate Unified Anomaly Segmentation
Wenxin MaFirst author
Participated works
Collaborative research as a contributing co-author.
2026
ECCV 2026
Concept-to-Pixel: Prompt-Free Universal Medical Image Segmentation
Combining semantic and geometric representations for medical segmentation without manual prompts.
WACV 2026
Equivariant Sampling for Improving Diffusion Model-based Image Restoration
Equivariant sampling improves diffusion-based image restoration without extra computational cost.
Patterns · 2026
A Self-Supervised Framework for Emphysema Anomaly Detection and Staging in Computed Tomography Scans
Self-supervised detection, localization, and staging of emphysema in CT scans.
News & milestones
[Mar 2026] Released Thinking with Spatial Code, exploring explicit 3D representations for video reasoning.
[Jan 2026] Released CausalSpatial, a benchmark for object-centric causal spatial reasoning.
[June 2025] 🎉 Two papers accepted at MICCAI 2025!
[May 2025] 🎉 A paper accepted at ACL 2025 (Findings)!
[Apr 2025] 🎤 Gave a talk at CSIG Wuhan Member Activity Center ‘Donghu Forum’ Frontier Paper Sharing Session!
[Feb 2025] 🎉 A paper accepted at CVPR 2025!
[Feb 2025] 🎉 A new preprint released on arXiv.
[Nov 2024] 🎉 Received China National Scholarship by Chinese Ministry of Education (Top 0.2%)!
[Jul 2024] 🎉 A paper accepted by WACV 2025 (oral)!
[Sep 2023] A New Start: My research life starts as a master student in MIRACLE lab! Good luck to myself!
Education & appointments
- PhD student, Computer Science — The University of Texas at Austin
Advised by Prof. Philipp Krähenbühl.
- Visiting Scholar — Whiting School of Engineering, Johns Hopkins University
Research with Prof. Alan Yuille and Ph.D. candidate Jieneng Chen at CCVL, on 3D spatial reasoning and AIGC.
- M.S., Biomedical Engineering — University of Science and Technology of China (USTC)
Advised by Prof. S. Kevin Zhou and Prof. Zihang Jiang, MIRACLE Lab. Weighted average score 90.7/100 (Top 5%).
- B.S., Biomedical Science — Zhiyuan Honored Program, Shanghai Jiao Tong University (SJTU)
Weighted average score 88.6/100.
Presentations & talks
[Apr 2025] Invited talk, “Frontier Paper Sharing Session” — CSIG Wuhan Member Activity Center, Donghu Forum
Teaching
- Teaching Assistant — Special Topic on Biomedical Engineering and Technological Innovation
- Teaching Assistant — Frontiers of Electronic Information
Professional service
Conference reviewer: ICLR · NeurIPS
Journal reviewer: Medical Image Analysis · IEEE Transactions on Multimedia (TMM)
Patent: A System for Unsupervised Anomaly Detection (since September 2023)
Honors & awards
[Nov 2024] 🎉 China National Scholarship by Chinese Ministry of Education (Top 0.2%)
[Oct 2024] 🎉 Outstanding Student Scholarship (Grade 1) by USTC (Top 30%)
[Oct 2023] 🎉 Outstanding Student Scholarship (Grade 1) by USTC (Top 30%)
[Nov 2022] 🎉 Zhiyuan Honored Scholarship by SJTU (Top 5%)
[Nov 2021] 🎉 Zhiyuan Honored Scholarship by SJTU (Top 5%)
[Nov 2020] 🎉 Zhiyuan Honored Scholarship by SJTU (Top 5%)
[Nov 2019] 🎉 Zhiyuan Honored Scholarship by SJTU (Top 5%)
Let’s connect
I welcome conversations about research and collaboration in computer vision, active exploration, and medical AI.