Paper | EF-SAI Dataset | Pre-trained Models | BibTex

Synthetic Aperture Imaging with Events and Frames

Wei Liao, Xiang Zhang, Lei Yu, Shijie Lin, Wen Yang, Ning Qiao
In CVPR'2022

The EF-SAI dataset is used for model training and evaluation. It can be download from Onedrive or Baidudrive

 python 3.6
 pytorch 1.7.1
 torchvision 0.8.2
 numpy, opencv, timm, matlpotlib, prefetch_generator

1. Clone this repo
2. Install PyTorch and dependencies from http://pytorch.org
3. Install required python packages

1. Download validation data from Onedrive or Baidudrive to ./EF-SAI-main/Val_data folder
2. Download pre-trained model from Onedrive or Baidudrive to ./EF-SAI-main/Pretrained folder

1. Change the parent directory to ./EF-SAI-main
2. Run testing script

python test.py 

3. The results will stored in the Results folder.