Tsung-Han Wu
Papers
1
Total Citations
21
H-Index
1
About
Tsung-Han Wu is a rising researcher in computer vision and 3D scene understanding, with a focus on depth estimation and sensor fusion. His most cited work, "S³: Learnable Sparse Signal Superdensity for Guided Depth Estimation" (2021, 21 citations), tackles a critical challenge in autonomous systems: improving dense depth maps from sparse, imbalanced sensor inputs like LiDAR or Radar. Wu introduces a learnable framework that generates a "superdense" signal from sparse guidance, significantly enhancing depth accuracy for robotics, 3D reconstruction, and augmented reality applications. This contribution addresses a fundamental bottleneck in real-world perception—how to effectively fuse sparse, irregular sensor data with dense image features—making his work highly relevant for autonomous driving and AR systems. By proposing a method that learns to densify and balance sparse signals, Wu advances the state of the art in guided depth estimation, offering a practical solution for scenarios where dense ground truth is unavailable. His research sits at the intersection of geometric computer vision and deep learning, with potential to impact both academic benchmarks and industry deployment.
Research Focus
Key Achievements
Top Papers
- 1