Awesome
Cross-Resolution-Flow-Propagation-for-Foveated-Video-Super-Resolution
Official implementation of Cross-Resolution Flow Propagation for Foveated Video Super-Resolution (CRFP) accepted by WACV 2023.
<img src="overview.png" width="600">Demo
Demonstration how CRFP deal with various value of $\sigma^T$ representing the noise induced by the movement of eye tracker during pupil detection. Note that regions beyond the foveated region are resolved under $8\times$ super-resolution.
$\sigma^T=10$
<img src="demo/sigma10.gif" width="600">$\sigma^T=50$
<img src="demo/sigma50.gif" width="600">$\sigma^T=100$
<img src="demo/sigma100.gif" width="600">Training and evaluation
To train the model, you need to install DCN first from https://github.com/jinfagang/DCNv2_latest
Run the following to start training
bash train.sh
To evaluate, run
bash eval.sh
To test, run
bash test.sh
References
Most of the code is referenced from