Learning to See the World Clearly and Comprehensively |
|
|
|
|
We aim to develop learning-based methods to reconstruct 4-D Light Fields (various perspectives), 3-D Hyperspectral data (rich spectrum information), and 3-D Point Clouds (geometry structure) from low-cost and convenient devices for truly perceiving the real-world, as well as the processing and analysis algorithms for such high-dimensional data in various applications and complex imaging environments. |
|
