Scholar
Xindi Wu
Google Scholar ID: hvnUnrUAAAAJ
Princeton University
Citations & Impact
Citations
2,546
H-index
11
i10-index
11
Publications
20
Co-authors
9
list available
Publications
Resume (English only)
Academic Achievements
- Paper 'Corgi: Cached Memory Guided Video Generation' accepted to WACV 2025; 'ConceptMix: A Compositional Image Generation Benchmark with Controllable Difficulty' accepted to NeurIPS D&B 2024; 'Vision-Language Dataset Distillation' won best paper award at ECCV 2024 Dataset Distillation Workshop; Developed and released projects like COMPACT, ICONS, etc.
Research Experience
- Gave a poster presentation at ICCV Workshop on Reliable and Interactable World Models 2025; Gave a talk at Datology AI Summer of Data Seminar 2025; Co-organized the ICCV 2025 Workshop on Curated Data for Efficient Learning; Interned at Nvidia Spatial Intelligence Lab during summer 2025; Led discussions or group talks, such as leading a discussion on Compositionality in Vision-Language Discussion Group in April 2025, etc.
Education
- Ph.D. student in Computer Science at Princeton University, advised by Prof. Olga Russakovsky; Master of Science in Computer Vision from Carnegie Mellon University, advised by Prof. Deva Ramanan and worked with Prof. Min Xu, Prof. Haohan Wang, and Prof. Kris Kitani; Admitted to the Honors Youth Program at Xi'an Jiaotong University.
Background
- Research interests: Data for efficient multimodal machine learning, focusing on tracing model performance back to training data composition and building more scalable and capable vision-language systems. Field: Computer Science.
Miscellany
- Welcomes students from underrepresented groups to reach out; Email, CV, Google Scholar, LinkedIn, and GitHub profiles are publicly available; Encourages questions and collaborations via email.
