pmoulon / sparse-ncnet

Implementation of Sparse-NCNet: "Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions"

Sparse Neighbourhood Consensus Networks

About

This is the implementation of the paper "Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions" by Ignacio Rocco, Relja Arandjelović and Josef Sivic [arXiv].

Installation

For installation instructions, please see INSTALL.md.

Quickstart

For a demo of the method, see the Jupyter notebook demo/demo.ipynb.

Training

To train a model with the default parameters run python train.py.

Evaluation on HPatches Sequences

1. Browse to eval/.
2. Run python eval_hpatches_extract.py adjusting the checkpoint and experiment name.
3. Use eval_hpatches_generate_plot.ipynb with the appropriate experiment name to generate the plot.

Evaluation on InLoc

In order to run the InLoc evaluation, you first need to clone the InLoc demo repo, and download and compile all the required depedencies. Then:

1. Browse to eval/.
2. Run python eval_inloc_extract.py adjusting the checkpoint and experiment name. This will generate a series of matches files in the datasets/inloc/matches/ folder that then need to be fed to the InLoc evaluation Matlab code.
3. Modify the eval/eval_inloc_compute_poses.m file provided in this repo to indicate the path of the InLoc demo repo, and the name of the experiment (the particular folder name inside datasets/inloc/matches/), and run it using Matlab.
4. Use the eval/eval_inloc_generate_plot.m file to plot the results from shortlist file generated in the previous stage: /your_path_to/InLoc_demo_old/experiment_name/shortlist_densePV.mat. Precomputed shortlist files are provided in datasets/inloc/shortlist.

Evaluation on Aachen Day-Night

In order to run the Aachen Day-Night evaluation, you first need to clone the Visualization benchmark repo, and download and compile all the required depedencies (in particular, you'll need to compile Colmap if you have not done so yet). Then:

1. Browse to eval/.
2. Run python eval_aachen_extract.py adjusting the checkpoint and experiment name.
3. Copy the eval_aachen_reconstruct.py file to visuallocalizationbenchmark/local_feature_evaluation and run it in the following way:

python eval_aachen_reconstruct.py 
	--dataset_path /path_to_aachen/aachen 
	--colmap_path /local/colmap/build/src/exe
	--method_name experiment_name

4. Upload the file /path_to_aachen/aachen/Aachen_eval_[experiment_name].txt to https://www.visuallocalization.net/ to get the results on this benchmark.

BibTeX

If you use this code in your project, please cite our paper:

@article{Rocco20,
        author       = "Rocco, I. and Arandjelovi\'c, R. and Sivic, J."
        title        = "Efficient Neighbourhood Consensus Networks via Submanifold Sparse Convolutions",
        journal      = "arXiv preprint arXiv:2004.10566",
        year         = "2020",
        }