Neuromorphic Imaging with Super-Resolution

@article{zhang2025tcsvt,
  title   = {Neuromorphic Imaging with Super-Resolution},
  author  = {Pei Zhang and Shuo Zhu and Chutian Wang and Yaping Zhao and Edmund Y. Lam},
  journal = {IEEE Transactions on Circuits and Systems for Video Technology},
  volume  = {35}, number = {2}, pages = {1715--1727},
  year    = {2025},
  doi     = {10.1109/TCSVT.2024.3482436}
}

Implementation

This repo provides an unoptimized prototype allowing you to make modifications as needed. You can try other networks and learning settings for various scenarios.

1. Prepare your event sample (with t, x, y, p entries) in the data folder.
2. Run

   CUDA_VISIBLE_DEVICES=0 python run_task.py
   

3. Check the result folder for output files.

Result

3D visualization (x, y, t) is intuitive to verify your result and its distribution: