Home

Awesome

(CVPR2022) TransMVSNet: Global Context-aware Multi-view Stereo Network with Transformers

Paper | Project Page | Arxiv | Models

Tips: If you meet any problems when reproduce our results, please contact Yikang Ding (dyk20@mails.tsinghua.edu.cn). We are happy to help you solve the problems and share our experience.

⚠ Change log

πŸ“” Introduction

In this paper, we present TransMVSNet, based on our exploration of feature matching in multi-view stereo (MVS). We analogize MVS back to its nature of a feature matching task and therefore propose a powerful Feature Matching Transformer (FMT) to leverage intra- (self-) and inter- (cross-) attention to aggregate long-range context information within and across images. To facilitate a better adaptation of the FMT, we leverage an Adaptive Receptive Field (ARF) module to ensure a smooth transit in scopes of features and bridge different stages with a feature pathway to pass transformed features and gradients across different scales. In addition, we apply pair-wise feature correlation to measure similarity between features, and adopt ambiguity-reducing focal loss to strengthen the supervision. To the best of our knowledge, TransMVSNet is the first attempt to leverage Transformer into the task of MVS. As a result, our method achieves state-of-the-art performance on DTU dataset, Tanks and Temples benchmark, and BlendedMVS dataset.

πŸ”§ Installation

Our code is tested with Python==3.6/3.7/3.8, PyTorch==1.6.0/1.7.0/1.9.0, CUDA==10.2 on Ubuntu-18.04 with NVIDIA GeForce RTX 2080Ti. Similar or higher version should work well.

To use TransMVSNet, clone this repo:

git clone https://github.com/MegviiRobot/TransMVSNet.git
cd TransMVSNet

We highly recommend using Anaconda to manage the python environment:

conda create -n transmvsnet python=3.6
conda activate transmvsnet
pip install -r requirements.txt
<!-- Install pytorch-fast-transformers by: ``` pip install --user pytorch-fast-transformers ``` -->

We also recommend using apex, you can install apex from the official repo.

πŸ“¦ Data preparation

In TransMVSNet, we mainly use DTU, BlendedMVS and Tanks and Temples to train and evaluate our models. You can prepare the corresponding data by following the instructions below.

βœ” DTU

For DTU training set, you can download the preprocessed DTU training data and Depths_raw (both from Original MVSNet), and unzip them to construct a dataset folder like:

dtu_training
 β”œβ”€β”€ Cameras
 β”œβ”€β”€ Depths
 β”œβ”€β”€ Depths_raw
 └── Rectified

For DTU testing set, you can download the preprocessed DTU testing data (from Original MVSNet) and unzip it as the test data folder, which should contain one cams folder, one images folder and one pair.txt file.

βœ” BlendedMVS

We use the low-res set of BlendedMVS dataset for both training and testing. You can download the low-res set from orignal BlendedMVS and unzip it to form the dataset folder like below:

BlendedMVS
 β”œβ”€β”€ 5a0271884e62597cdee0d0eb
 β”‚     β”œβ”€β”€ blended_images
 β”‚     β”œβ”€β”€ cams
 β”‚     └── rendered_depth_maps
 β”œβ”€β”€ 59338e76772c3e6384afbb15
 β”œβ”€β”€ 59f363a8b45be22330016cad
 β”œβ”€β”€ ...
 β”œβ”€β”€ all_list.txt
 β”œβ”€β”€ training_list.txt
 └── validation_list.txt

βœ” Tanks and Temples

Download our preprocessed Tanks and Temples dataset and unzip it to form the dataset folder like below:

tankandtemples
 β”œβ”€β”€ advanced
 β”‚  β”œβ”€β”€ Auditorium
 β”‚  β”œβ”€β”€ Ballroom
 β”‚  β”œβ”€β”€ ...
 β”‚  └── Temple
 └── intermediate
        β”œβ”€β”€ Family
        β”œβ”€β”€ Francis
        β”œβ”€β”€ ...
        └── Train

πŸ“ˆ Training

βœ” Training on DTU

Set the configuration in scripts/train.sh:

To train your own model, just run:

bash scripts/train.sh

You can conveniently modify more hyper-parameters in scripts/train.sh according to the argparser in train.py, such as summary_freq, save_freq, and so on.

βœ” Finetune on BlendedMVS

For a fair comparison with other SOTA methods on Tanks and Temples benchmark, we finetune our model on BlendedMVS dataset after training on DTU dataset.

Set the configuration in scripts/train_bld_fintune.sh:

To finetune your own model, just run:

bash scripts/train_bld_fintune.sh

πŸ“Š Testing

For easy testing, you can download our pre-trained models and put them in checkpoints folder, or use your own models and follow the instruction below.

βœ” Testing on DTU

Important Tips: to reproduce our reported results, you need to:

To start testing, set the configuration in scripts/test_dtu.sh:

Run:

bash scripts/test_dtu.sh

Note: You can use the gipuma fusion method or normal fusion method to fuse the point clouds. In our experiments, we use the gipuma fusion method by default. With using the uploaded ckpt and latest code, these two fusion methods would get the below results:

FuseOverall
gipuma0.304
normal0.314
<!-- The simple instruction for installing and compiling `gipuma` can be found [here](https://github.com/YoYo000/MVSNet#post-processing). The installed gipuma is a modified version from [Yao Yao](https://github.com/YoYo000/fusibile).-->

To install the gipuma, clone the modified version from Yao Yao. Modify the line-10 in CMakeLists.txt to suit your GPUs. Othervise you would meet warnings when compile it, which would lead to failure and get 0 points in fused point cloud. For example, if you use 2080Ti GPU, modify the line-10 to:

set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS};-O3 --use_fast_math --ptxas-options=-v -std=c++11 --compiler-options -Wall -gencode arch=compute_70,code=sm_70)

If you use other kind of GPUs, please modify the arch code to suit your device (arch=compute_XX,code=sm_XX). Then install it by cmake . and make, which will generate the executable file at FUSIBILE_EXE_PATH. Please note

For quantitative evaluation on DTU dataset, download SampleSet and Points. Unzip them and place Points folder in SampleSet/MVS Data/. The structure looks like:

SampleSet
β”œβ”€β”€MVS Data
      └──Points

In DTU-MATLAB/BaseEvalMain_web.m, set dataPath as path to SampleSet/MVS Data/, plyPath as directory that stores the reconstructed point clouds and resultsPath as directory to store the evaluation results. Then run DTU-MATLAB/BaseEvalMain_web.m in matlab.

We also upload our final point cloud results to here. You can easily download them and evaluate them using the MATLAB scripts, the results look like:

Acc. (mm)Comp. (mm)Overall (mm)
0.3210.2890.305

βœ” Testing on Tanks and Temples

We recommend using the finetuned models (model_bld.ckpt) to test on Tanks and Temples benchmark.

Similarly, set the configuration in scripts/test_tnt.sh:

To generate point cloud results, just run:

bash scripts/test_tnt.sh

Note that:

For quantitative evaluation, you can upload your point clouds to Tanks and Temples benchmark.

πŸ”— Citation

@inproceedings{ding2022transmvsnet,
  title={Transmvsnet: Global context-aware multi-view stereo network with transformers},
  author={Ding, Yikang and Yuan, Wentao and Zhu, Qingtian and Zhang, Haotian and Liu, Xiangyue and Wang, Yuanjiang and Liu, Xiao},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={8585--8594},
  year={2022}
}

πŸ“Œ Acknowledgments

We borrow some code from CasMVSNet, LoFTR and AA-RMVSNet. We thank the authors for releasing the source code.