@inproceedings{zhang2026paprupclose,
  title={PAPR Up-close: Close-up Neural Point Rendering without Holes},
  author={Yanshu Zhang and Chirag Vashist and Shichong Peng and Ke Li},
  booktitle={International Conference on 3D Vision},
  year={2026}
}

TL;DRHole-free close-up neural point rendering
We extend PAPR for robust close-up neural point rendering and significantly reduce holes and artifacts while preserving fine details.

@inproceedings{aghabozorgi2026wimle,
  title={{WIMLE}: Uncertainty-Aware World Models with {IMLE} for Sample-Efficient Continuous Control},
  author={Mehran Aghabozorgi and Alireza Moazeni and Yanshu Zhang and Ke Li},
  booktitle={The Fourteenth International Conference on Learning Representations},
  year={2026}
}

TL;DRMulti-modal world models with uncertainty-aware policy learning for model-based RL
WIMLE learns stochastic, multi-modal world models using IMLE and weights synthetic transitions by predictive confidence, achieving state-of-the-art sample efficiency across 40 continuous-control tasks in DeepMind Control, HumanoidBench, and MyoSuite.

@inproceedings{peng2024papr,
  title={PAPR in Motion: Seamless Point-level 3D Scene Interpolation},
  author={Shichong Peng and Yanshu Zhang and Ke Li},
  booktitle={IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  year={2024}
}

TL;DRSeamless point-level 4D motion interpolation
We introduce the novel problem of point-level 3D scene interpolation. Given observations of a scene at two distinct states from multiple views, the goal is to synthesize a smooth point-level interpolation between them, without any intermediate supervision. Our method, PAPR in Motion, builds upon Proximity Attention Point Rendering (PAPR) technique, and generates seamless interpolations of both the scene geometry and appearance.

@article{moazeni2024intrinsic,
  title={Intrinsic PAPR for Point-level 3D Scene Albedo and Shading Editing},
  author={Alireza Moazeni and Shichong Peng and Ke Li},
  journal={arXiv preprint arXiv:2407.00500},
  year={2024}
}

TL;DRPoint-level 3D albedo and shading editing via intrinsic decomposition
Intrinsic PAPR decomposes a 3D scene into albedo and shading components using point-based neural rendering, enabling detailed point-level editing that remains consistent across viewpoints without relying on complex shading models or simplistic priors.

@inproceedings{zhang2023papr,
  title={PAPR: Proximity Attention Point Rendering},
  author={Yanshu Zhang and Shichong Peng and Alireza Moazeni and Ke Li},
  booktitle={Thirty-seventh Conference on Neural Information Processing Systems},
  year={2023}
}

TL;DRReconstruct and render point clouds using attention
PAPR is a point-based surface representation that uses proximity attention to interpolate between nearby points to rays for rendering high-quality images, enabling non-volume-preserving geometry deformation by directly adjusting point positions, and, unlike 3D Gaussian Splatting, it avoids creating holes while preserving texture details after deformation.