matteocali/nlos-imaging-from-itof-data

Implementation of the project in collaboration with Sony Europe B.V. that aims to perform imaging of a 3D object located in Non-Line-of-Sight (NLoS) using an indirect Time of Flight (iToF) in combiantion with a Neural Network (NN)

NIGHT - Non-Line-of-Sight Imaging from Indirect Time of Flight Data

This repository contains all the code and tools related to the work pressented in the paper NIGHT - Non-Line-of-Sight Imaging from Indirect Time of Flight Data.

The repository is organized in three sub folder:

1. additional_scripts: contains all the code to genrate and pre-proces the presented synthetic dataset;
2. itofNet_code: contains the code of the proposed network;
3. tools: contains all the additional data and tools used in the previous two sections.

To download the dataset please use the following links:

• synthetic dataset available here,
• splits used in the paper are available here.

How to run the code

In order to run the proposed code follow the instruction presented in the additional_scripts and itofNet_code README.

Authors

• Matteo Caligiuri,
• Adriano Simonetto,
• Pietro Zanuttigh.

Cite as

@article{caligiuri2024night,
  title={NIGHT--Non-Line-of-Sight Imaging from Indirect Time of Flight Data},
  author={Caligiuri, Matteo and Simonetto, Adriano and Zanuttigh, Pietro},
  journal={arXiv preprint arXiv:2403.19376},
  year={2024}
}