Hao Su
Hao Su co-authored PointNet, first submitted in December 2016 and presented at CVPR 2017. The network processes point sets directly for classification and segmentation while accounting for point-order invariance. Su joined UCSD’s faculty on 1 July 2017 and worked on ShapeNet and 3D deep learning. His associate professorship at UCSD has ended.
- Works
- 85
Publication numbers from OpenAlex, as of 14 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.
Timeline
-
2016–2017
PointNet
Su co-authored PointNet with Charles Qi, Kaichun Mo, and Leonidas Guibas. The manuscript was first submitted in December 2016 and appeared at CVPR 2017. It introduced a neural network that directly processes point sets, with applications in object classification, part segmentation, and scene semantic parsing.
-
2016–2017
Handling unordered 3D data
PointNet accounts for permutation invariance: reordering the input points should preserve their meaning as a set. The paper also analyses responses to input perturbation and corruption. These methods address the structure of point-cloud data.
-
2017
Joining UC San Diego
Su joined UCSD’s computer science department on 1 July 2017. His research covered vision, graphics, learning and robotics, including work on ShapeNet.
-
As of October 2026
Former UCSD associate professorship
Su previously held an associate professorship at UCSD.
Sources and credits
Photo: cse.ucsd.edu; creator not stated. License: Not stated by source. Source: cse.ucsd.edu. Cropped to a square and resized. Ask us to remove or replace this photo.