Computer Graphics, Computer Vision, Artificial Intelligence, Computational Optics

Dr. HUO’s research covers a wide range of areas including realistic, real-time, and neural rendering, image retrieval, interpolation, segmentation, 3D sensing, and reconstruction, and optical neural networks. He developed the first incoherent optoelectronic neural network, based on theories and methods related to compressed sensing, parallel computing, deep learning, and reinforcement learning. 

Representative Publications in the Past 2 Years:

1. Adaptive Incident Radiance Field Sampling and Reconstruction Using Deep Reinforcement Learning. ACM Transaction on TOG (selected as journal cover).

2. Weakly-Supervised Contrastive Learning in Path Manifold for Monte Carlo Image Reconstruction. ACM Transactions on Graphics.

3. PowerNet: Learning-Based Real-Time Power-Budget Rendering. IEEE Transactions on Visualization and Computer Graphics.

4. Single Image Reflection Removal with Physically-Based Training Images. CVPR (oral presentation).

5. Hypergraph Propagation and Community Selection for Objects Retrieval. NeurIPS.