Juan Felipe Pérez-Juste Abascal, Ph.D.

Deep learning for image reconstruction

​

​

SPyRiT  

Repository: 

https://github.com/openspyrit/spyrit

​

Contribute to SPyRiT, a PyTorch-based toolbox for deep image reconstruction.

Image reconstruction: tutorials, colab

​​

​

​

ResPr-UNet-3D-Denoising-Efficient-Pipeline-TF-keras  

​

​

Repository: 

https://github.com/jabascal/ResPr-UNet-3D-Denoising-Efficient-Pipeline-TF-keras/

​

Modified U-Net architecture (ResPrU-Net) that exploits a Prior Image for 3D image denoising. It extends sparse, prior-based methods to learning approaches.

​

Efficient data pipelines for 3D denoising.

​

Comparison of different architectures (CNN, U-Net, ResPr-UNet), sparse based methods (TV, BM3D) and 2D and 3D denoising.

​

Assessed on kits19 challenge dataset.

​

Code and results for the paper A residual U-Net network with image prior for 3D image denoising, Proc. Eur. Signal Process. Conf. EUSIPCO, pp. 1264-1268, 2020. (hal-02500664)

​