emjian / edncnn

Source code for EDnCNN paper from CVPR 2020

Event Probability Mask (EPM) and Event Denoising Convolutional Neural Network (EDnCNN) for Neuromorphic Cameras

Contains a MATLAB implementation for our paper. If you find this code useful in your research, please consider citing:

@article{baldwin2020event,
  title={Event Probability Mask (EPM) and Event Denoising Convolutional Neural Network (EDnCNN) for Neuromorphic Cameras},
  author={Baldwin, R and Almatrafi, Mohammed and Asari, Vijayan and Hirakawa, Keigo},
  journal={arXiv preprint arXiv:2003.08282},
  year={2020}
}

This code was tested on an Ubuntu 18.04 system (i7-8700 CPU, 64GB RAM, and GeForce RTX 2080Ti GPU) running MATLAB 2019b. Matlab's image processing and computer vision toolboxes are required.

DVSNOISE20 Dataset (optional)

To download the dataset use: dvsnoise20.

Reading AEDAT data into MATLAB (optional)

To read AEDAT (jAER) data into MATLAB use: AedatTools or AedatTools Alt.

Reading AEDAT4 data into MATLAB (optional)

To read AEDAT4 (DV) data into MATLAB use: aedat4tomat.