Simlearner3D is a deep learning library that allows to learn similarity cues for large scale 3D reconstruction from aerial and satellite imagery.
It is built to prepare, structure training/evaluation/test datasets, train a variety of neural networks and test the performance of learned models.
Simlearner3d is built upon PyTorch.
Its structure was bootstraped from this code template, which heavily relies on Hydra and Pytorch-Lightning to enable flexible and rapid iterations of deep learning experiments.
→ For installation and usage, please refer to Documentation.
Please cite simlearner3d if it helped your own research. Here is an example BibTex entry:
@INPROCEEDINGS{10208916,
author={Chebbi, Mohamed Ali and Rupnik, Ewelina and Pierrot-Deseilligny, Marc and Lopes, Paul},
booktitle={2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)},
title={DeepSim-Nets: Deep Similarity Networks for Stereo Image Matching},
year={2023},
pages={2097-2105}}