Related Experiment Video
Updated: Oct 12, 2025
06:25
Time Multiplexing Super Resolving Technique for Imaging from a Moving Platform
Published on: February 12, 2014
8.6K
Passive Non-Line-of-Sight Imaging Using Optimal Transport
Summary
This study introduces NLOS-OT, a new framework for passive non-line-of-sight (NLOS) imaging, which reconstructs complex hidden scenes with high quality. A large-scale dataset, NLOS-Passive, was also created to advance passive NLOS imaging research.
Area of Science:
- Computer Vision
- Computational Imaging
- Machine Learning
Background:
- Passive non-line-of-sight (NLOS) imaging is a rapidly developing field.
- Existing methods struggle with complex scenes, low-quality reconstructions, and limited datasets.
Purpose of the Study:
- To develop a novel framework for high-quality passive NLOS imaging of complex scenes.
- To introduce the first large-scale dataset for evaluating passive NLOS imaging algorithms.
Main Methods:
- NLOS-OT framework utilizes manifold embedding and optimal transport.
- Transforms high-dimensional reconstruction into low-dimensional manifold mapping.
- Introduces the NLOS-Passive dataset with over 3.2 million images.
Main Results:
- NLOS-OT significantly improves reconstruction quality for complex hidden scenes.
- The NLOS-OT framework outperforms state-of-the-art methods on the NLOS-Passive dataset.
- Demonstrates superior performance in passive NLOS imaging tasks.
Conclusions:
- NLOS-OT offers a robust solution for challenging passive NLOS imaging problems.
- The NLOS-Passive dataset provides a valuable resource for future research.
- This work represents a significant advancement in learning-based passive NLOS imaging.
