CAD Generative AI
Parametric and editable 3D model generation from text, point clouds, and other multimodal inputs.
- NURBGen AAAI 2026
- TextCAD NeurIPS 2024 Spotlight
- CAD-SIGNet CVPR 2024 Highlight
About
I am a Research Scientist at Spectral Labs working on foundation model for spatial intelligence, and 3D designing.
Parametric and editable 3D model generation from text, point clouds, and other multimodal inputs.
High-fidelity 3D models and scenes from text, images, and video, with an emphasis on scalable representations.
Geometric-semantic alignment for AI systems that can reason about objects, structure, and editable scenes.
Compact surface and mesh representations for high-resolution generation without sacrificing editability.
Reviewer
Teaching
Mentored Students