Discrete time convolution for fast event-based stereo

This code is a demo of our CVPR 2022 paper "Discrete time convolution for fast event-based stereo".

dataset

Change the MVSEC/DSEC dataset address in dataset.py.

Data preprocess for DSEC:
Follow from https://github.com/uzh-rpg/DSEC.

Data process for MVSEC (SBT):
Step 1: From h5 file to npy file: follow from https://github.com/tlkvstepan/event_stereo_ICCV2019.
Step 2: Using SBT method to proccess npy file: python code_of_mvsec/data_preprocess.py

cofe for mvsec

• code_of_mvsec/DTC_pds_for_mvsec
  usage: DTC-pds

• code_of_mvsec/DTC_SPADE_for_mvsec
  usage: DTC-spade

For training/test procedure, just execute:
bash train.sh/test.sh

code for Dsec

• code_of_Dsec/LTC_Dsec
  usage: generate DSEC website test result by DTC-PDS

• code_of_Dsec/LTC_Dsec_spade
  usage: generate DSEC website test result by DTC-SPADE

• code_of_Dsec/LTC_for_Dsec/LTC_Dsec_clear_version
  usage: DTC-PDS

• code_of_Dsec/LTC_for_Dsec/LTC_Dsec_spade_clear_version
  usage: DTC-SPADE

For training/test procedure, just execute:
bash train.sh/test.sh

Paper Reference

@inproceedings{zhang2022discrete,
title={Discrete Time Convolution for Fast Event-Based Stereo},
author={Zhang, Kaixuan and Che, Kaiwei and Zhang, Jianguo and Cheng, Jie and Zhang, Ziyang and Guo, Qinghai and Leng, Luziwei},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={8676--8686},
year={2022}
}

Our code is developed based on the code from ICCV2019 paper "Learning an event sequence embedding for dense event-based deep stereo"

paper: https://openaccess.thecvf.com/content_ICCV_2019/papers/Tulyakov_Learning_an_Event_Sequence_Embedding_for_Dense_Event-Based_Deep_Stereo_ICCV_2019_paper.pdf
code: https://github.com/tlkvstepan/event_stereo_ICCV2019

Liscence

This open-source project is not an official Huawei product, and Huawei is not expected to provide support for this project.