Submodule used in hloc toolbox
Our method first inferred the latent topics (high-level context information) for each image and then use them to explicitly learn robust feature representation for the matching task. Please check out the details in our paper
Overall Architecture:
- Release training and evaluation code on MegaDepth and ScanNet
- Evaluation on HPatches, Aachen Day&Night, and InLoc
- Evaluation for Image Matching Challenge
All experiments in this paper are implemented on the Ubuntu environment with a NVIDIA driver of at least 430.64 and CUDA 10.1.
First, create a virtual environment by anaconda as follows,
conda create -n topicfm python=3.8
conda activate topicfm
conda install pytorch==1.8.1 torchvision==0.9.1 cudatoolkit=10.1 -c pytorch
pip install -r requirements.txt
# using pip to install any missing packages
The proposed method is trained on the MegaDepth dataset and evaluated on the MegaDepth test, ScanNet, HPatches, Aachen Day and Night (v1.1), and InLoc dataset. All these datasets are large, so we cannot include them in this code. The following descriptions help download these datasets.
This dataset is used for both training and evaluation (Li and Snavely 2018). To use this dataset with our code, please follow the instruction of LoFTR (Sun et al. 2021)
We only use 1500 image pairs of ScanNet (Dai et al. 2017) for evaluation. Please download and prepare test data of ScanNet provided by LoFTR.
To train our model, we recommend to use GPUs card as much as possible, and each GPU should be at least 12GB.
In our settings, we train on 4 GPUs, each of which is 12GB.
Please setup your hardware environment in scripts/reproduce_train/outdoor.sh
.
And then run this command to start training.
bash scripts/reproduce_train/outdoor.sh
We then provide the trained model in pretrained/model_best.ckpt
bash scripts/reproduce_test/outdoor.sh
bash scripts/reproduce_test/indoor.sh
To evaluate on these datasets, we integrate our code to the image-matching-toolbox provided by Zhou et al. (2021). The updated code is available here. After cloning this code, please follow instructions of image-matching-toolbox to install all required packages and prepare data for evaluation.
Then, run these commands to perform evaluation: (note that all hyperparameter settings are in configs/topicfm.yml
)
HPatches (homography estimation)
python -m immatch.eval_hpatches --gpu 0 --config 'topicfm' --task 'both' --h_solver 'cv' --ransac_thres 3 --root_dir . --odir 'outputs/hpatches'
Aachen Day-Night v1.1 (visual localization)
python -m immatch.eval_aachen --gpu 0 --config 'topicfm' --colmap <path to use colmap> --benchmark_name 'aachen_v1.1'
InLoc (visual localization)
python -m immatch.eval_inloc --gpu 0 --config 'topicfm'
IMC-2022 was held on Kaggle. Most high ranking methods were achieved by using an ensemble method which combines the matching results of various state-of-the-art methods including LoFTR, SuperPoint+SuperGlue, MatchFormer, or QuadTree Attention.
In this evaluation, we only submit the results produced by our method (TopicFM) alone. Please refer to this notebook. This table compares our results with the other methods such as LoFTR (ref. here), SP+SuperGlue (ref. here).
Public Score | Private Score | |
---|---|---|
SP + SuperGlue | 0.678 | 0.677 |
LoFTR | 0.726 | 0.736 |
TopicFM (ours) | 0.804 | 0.811 |
The runtime reported in the paper is measured by averaging runtime of 1500 image pairs of the ScanNet evaluation dataset.
The image size can be changed at configs/data/scannet_test_1500.py
python visualization.py --method <method_name> --dataset_name "scannet" --measure_time --no_viz
# note that method_name is in ["topicfm", "loftr"]
To measure time for LoFTR, please download the LoFTR's code as follows:
git submodule update --init
# download pretrained models
mkdir third_party/loftr/pretrained
gdown --id 1M-VD35-qdB5Iw-AtbDBCKC7hPolFW9UY -O third_party/loftr/pretrained/outdoor_ds.ckpt
If you find this work useful, please cite this:
@article{giang2022topicfm,
title={TopicFM: Robust and Interpretable Topic-assisted Feature Matching},
author={Giang, Khang Truong and Song, Soohwan and Jo, Sungho},
journal={arXiv preprint arXiv:2207.00328},
year={2022}
}
This code is built based on LoFTR. We thank the authors for their useful source code.