dvs-voltmeter's Introduction

DVS-Voltmeter

Code repo for the paper 'DVS-Voltmeter: Stochastic Process-based Event Simulator for Dynamic Vision Sensors'.

Prerequisites

easydict == 1.9
pytorch >= 1.8
numpy >= 1.20.1
opencv-python == 4.5.1.48
tqdm == 4.49.0

The code may be compatible with lower versions, while the aforementioned ones have been tested.

Dataset

The sample input video frames to try DVS-Voltmeter with are in samples on google drive. Download samples and put them in data_samples/interp folder.

To simulate events from other videos, put high frame-rate videos (can be obtained by video interpolation) in data_samples/interp folder and modify the data index tree as following:

├── [data_samples]
│   ├── interp
│   │   ├── videoname1
│   │   │   ├── info.txt
│   │   │   ├── framename11.png
│   │   │   ├── framename12.png
│   │   ├── videoname2
│   │   │   ├── info.txt
│   │   │   ├── framename21.png
│   │   │   ├── framename22.png

The video info file info.txt records the path and timestamp ($\mu s$) of each frame.

Usage

Configure the src/config.py file. For detailed configuration, please refer to the config file.
Run python main.py

Biblography

If you find our work useful, please use the following citation.

@inproceedings{lin2022dvsvoltmeter,
  title={DVS-Voltmeter: Stochastic Process-based Event Simulator for Dynamic Vision Sensors},
  author={Lin, Songnan and Ma, Ye and Guo, Zhenhua and Wen, Bihan},
  booktitle={ECCV},
  year={2022}
}