# Deep Non-Line-of-Sight Reconstruction

arXiv: https://arxiv.org/abs/2001.09067 · HTML: https://mlsanity.com/paper/2001.09067/

## Key facts

- Cited by 21 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Jan 2020.

## Built on by (most cited first, 21 of 21)

- [Passive Inter-Photon Imaging](https://mlsanity.com/paper/2104.00059/) (2021)
- [Physics to the Rescue: Deep Non-line-of-sight Reconstruction for High-speed Imaging](https://mlsanity.com/paper/2205.01679/) (2022)
- [Recent Advances on Non-Line-of-Sight Imaging: Conventional Physical Models, Deep Learning, and New Scenes](https://mlsanity.com/paper/2104.13807/) (2021)
- [Few-shot Non-line-of-sight Imaging with Signal-surface Collaborative Regularization](https://mlsanity.com/paper/2211.15367/) (2022)
- [Self-Calibrating, Fully Differentiable NLOS Inverse Rendering](https://mlsanity.com/paper/2309.12047/) (2023)
- [Learning to Enhance Aperture Phasor Field for Non-Line-of-Sight Imaging](https://mlsanity.com/paper/2407.18574/) (2024)
- [Occlusion Fields: An Implicit Representation for Non-Line-of-Sight Surface Reconstruction](https://mlsanity.com/paper/2203.08657/) (2022)
- [Curvature regularization for Non-line-of-sight Imaging from Under-sampled Data](https://mlsanity.com/paper/2301.00406/) (2023)
- [Generalizable Non-Line-of-Sight Imaging with Learnable Physical Priors](https://mlsanity.com/paper/2409.14011/) (2024)
- [Towards Non-Line-of-Sight Photography](https://mlsanity.com/paper/2109.07783/) (2021)
- [Domain Reduction Strategy for Non Line of Sight Imaging](https://mlsanity.com/paper/2308.10269/) (2023)
- [MARMOT: Masked Autoencoder for Modeling Transient Imaging](https://mlsanity.com/paper/2506.08470/) (2025)
- [A comprehensive study of time-of-flight non-line-of-sight imaging](https://mlsanity.com/paper/2603.09548/) (2026)
- [Fast Non-line-of-sight Imaging with Two-step Deep Remapping](https://mlsanity.com/paper/2101.10492/) (2021)
- [NIGHT -- Non-Line-of-Sight Imaging from Indirect Time of Flight Data](https://mlsanity.com/paper/2403.19376/) (2024)
- [Dual-branch Graph Feature Learning for NLOS Imaging](https://mlsanity.com/paper/2502.19683/) (2025)
- [TransiT: Transient Transformer for Non-line-of-sight Videography](https://mlsanity.com/paper/2503.11328/) (2025)
- [Fast and Memory-efficient Non-line-of-sight Imaging with Quasi-Fresnel Transform](https://mlsanity.com/paper/2508.02003/) (2025)
- [3D Reconstruction from Transient Measurements with Time-Resolved Transformer](https://mlsanity.com/paper/2510.09205/) (2025)
- [SuperEx: Enhancing Indoor Mapping and Exploration using Non-Line-of-Sight Perception](https://mlsanity.com/paper/2510.10506/) (2025)
- [Cascaded Non-Line-of-Sight Imaging](https://mlsanity.com/paper/2609.16017/) (2026)

## Abstract (arXiv)

The recent years have seen a surge of interest in methods for imaging beyond the direct line of sight. The most prominent techniques rely on time-resolved optical impulse responses, obtained by illuminating a diffuse wall with an ultrashort light pulse and observing multi-bounce indirect reflections with an ultrafast time-resolved imager. Reconstruction of geometry from such data, however, is a complex non-linear inverse problem that comes with substantial computational demands. In this paper, we employ convolutional feed-forward networks for solving the reconstruction problem efficiently while maintaining good reconstruction quality. Specifically, we devise a tailored autoencoder architecture, trained end-to-end, that maps transient images directly to a depth map representation. Training is done using an efficient transient renderer for diffuse three-bounce indirect light transport that enables the quick generation of large amounts of training data for the network. We examine the performance of our method on a variety of synthetic and experimental datasets and its dependency on the choice of training data and augmentation strategies, as well as architectural features. We demonstrate that our feed-forward network, even though it is trained solely on synthetic data, generalizes to measured data from SPAD sensors and is able to obtain results that are competitive with model-based reconstruction methods.