I am a researcher/developer focused on Spatial Perception and Foundation Models. My work aims to bridge the gap between multimodal learning and 3D geometric understanding.
- 🔭 Focus: Spatial Perception and Understanding.
- 🌱 Learning: Flow-matching, 3D Gaussian Splatting, Advanced 3D Representations.
- 👯 Collaborate: Computer Vision, Multimodal LLMs (MLLM), Foundation Models.
- 💬 Ask me about: CV, PEFT, or 3D geometry in MLLMs.
|
Orient-Anything-V2 (NeurIPS 2025 Spotlight) 
|
Orient-Anything (ICML 2025) 
|
| Project | Conference | Highlights |
|---|---|---|
| Orient-Anything-V2 | NeurIPS 2025 Spotlight | Unifying Orientation and Rotation Understanding |
| Orient-Anything | ICML 2025 | Robust Object Orientation Estimation |
| DSI-Bench | arXiv 2025 | A Benchmark for Spatial Intelligence |
| OmniBind | ICLR 2025 | Multi-modal Binding Foundation Models |
| FreeBind | ICML 2024 | Flexible Modality Alignment |
| Ex-MCR | NeurIPS 2024 | Efficient Multi-modal Learning & PEFT |