Bridging Stereo Matching and Optical Flow via Spatiotemporal Correspondence, CVPR 2019
PyTorch implementaton of the following paper. In this paper, we propose a unified model for unsupervised stereo matching and optical flow estimation using a single neural network.
Bridging Stereo Matching and Optical Flow via Spatiotemporal Correspondence
Hsueh-Ying Lai, Yi-Hsuan Tsai, Wei-Chen Chiu
IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019.
Please cite our paper if you find it useful for your research. [Project Page]
@inproceedings{lai19cvpr,
title = {Bridging Stereo Matching and Optical Flow via Spatiotemporal Correspondence},
author = {Hsueh-Ying Lai and Yi-Hsuan Tsai and Wei-Chen Chiu},
booktitle = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
year = {2019}
}
wget -i utils/kitti_archives_to_download.txt -P ~/dataset/
KITTI_PATH/
|--training/
|--testing/
git clone https://github.com/lelimite4444/BridgeDepthFlow
cd BridgeDepthFlow
We use kitti split as example.
python train.py --data_path ~/dataset/
--filenames_file ./utils/filenames/kitti_train_files_png_4frames.txt
--checkpoint_path YOUR_CHECKPOINT_PATH
The chosen --type_of_2warp
from 0 ~ 2 correponds to three types of different 2warp function in Figure 4 of our paper.
The --model_name
flag allows you to choose which model you want to train on. We provide the PyTorch version of both monodepth and PWC-Net.
We use the validation set of KITTI 2015 as example. The ground truth of optical flow includes occluded area.
python test_flow.py --data_path KITTI_PATH
--filenames_file ./utils/filenames/kitti_flow_val_files_occ_200.txt
--checkpoint_path YOUR_CHECKPOINT_PATH/TRAINED_MODEL_NAME
python test_stereo.py --data_path KITTI_PATH
--filenames_file ./utils/filenames/kitti_stereo_2015_test_files.txt
--checkpoint_path YOUR_CHECKPOINT_PATH/TRAINED_MODEL_NAME
The network will output disparities.npy
, containing all the estimated disparities of test data. You need to evaluate it by running:
python utils/evaluate_kitti.py --split kitti --predicted_disp_path ./disparities.npy --gt_path ~/dataset/
Resample2d
and custom layers Correlation
which PWC-Net relys on are implemented by NVIDIA-flownet2