Chris Choy 3D vision researcher
My research moves between sparse computation, geometric learning, and open-vocabulary perception—from the operators underneath a model to the objects it can name in a scene.
Research map 01—03
Libraries, neural operators, and architectures that connect efficient 3D computation to scene understanding.
Neural network libraries sparse computation 4D perception
Connect language to complete 3D instances without turning perception into a slow, multi-stage pipeline.
3D-language data instance masks proposal-free decoding
Learn correspondence and pose while keeping the structure of rotation, distance, and registration explicit.
Correspondence rotation registration
Selected work 03
Selected across systems, perception, and geometry.
Imagine asking a robot: “bring me the Snoopy plush from the corner of the room.” For this to work, the robot needs to find the Snoopy in 3D space — a category it has likely never seen during...
Read entry
In many robotics and VR/AR applications, 3D-videos are readily-available sources of input (a continuous sequence of depth images, or LIDAR scans). However, those 3D-videos are processed frame-by-frame either through 2D convnets or 3D perception algorithms. In this work,...
Read entry
Predicting 3D rotations from neural networks sounds straightforward—output a rotation matrix and move on. But rotations live on $\mathrm{SO}(3)$, a curved 3-dimensional manifold inside $\mathbb{R}^{3 \times 3}$ with the constraints $R^\top R = I$ and $\det(R) = 1$....
Read entry
Notebook Latest entries
research
research
research
programming
programming
publications