CVPR 2025: 'SPAR3D: Stable Point-Aware Reconstruction of 3D Objects from Single Images' – a fast single-image 3D reconstructor enabling interactive edits.
CVPR 2025: Co-first author of 'SF3D: Stable Fast 3D Mesh Reconstruction with UV-unwrapping and Illumination Disentanglement', generating textured UV-unwrapped 3D models in under a second.
CVPR 2025 (Highlight): Contributed to 'Symmetry Strikes Back: From Single-Image Symmetry Detection to 3D Generation', a zero-shot 3D symmetry detector.
Tech Report 2024: Co-developed TripoSR, a large open-source model for high-quality single-image 3D reconstruction under one second.
CVPR 2024: 'PointInfinity: Resolution-Invariant Point Diffusion Models' – a point diffusion model with resolution-invariant generation.
CVPR 2024: First author of 'ZeroShape: Regression-based Zero-shot Shape Reconstruction', achieving SOTA with high efficiency and low data budget.
CVPR 2023: Published 'ShapeClipper: Scalable 3D Shape Learning from Single-View Images via Geometric and CLIP-based Consistency'.
NeurIPS 2023 (Datasets and Benchmarks Track): Contributed to 'Low-shot Object Learning with Mutual Exclusivity Bias'.
NeurIPS 2022: Co-authored 'Learning Dense Object Descriptors from Multiple Views for Low-shot Category Generalization'.
ECCV 2022: First author of 'Planes vs. Chairs: Category-guided 3D shape learning without any 3D cues', a 3D-unsupervised multi-category shape learner.