Code for our IEEE TCSVT Paper: Lightweight Modules for Efficient Deep Learning based Image Restoration
Authors: Avisek Lahiri*, Sourav Bairagya*, Sutanu Bera, Siddhant Haldar, Prabir Kumar Biswas
(* equal contribution)
- Paper Link: https://arxiv.org/abs/2007.05835
- IEEE Early Access Link: https://ieeexplore.ieee.org/document/9134805
- Paper provides re-usable modules to be plugged and played to compress a given CNN
- Select any favourite full-scale baseline for low-level vision applications
- Replace 3X3 conv by LIST layer
- Replace dilated conv layer GSAT layer
- Achieve efficient up/down-sample with Bilinear SubSampling followed by LIST layer
This contains the basic proposed modules in Tensorflow
TensorflowExamples/basicModules.py contains the proposed LIST, GSAT modules
It also contains the framework for LIST based up/down-sampling in a CNN
It contains the basic proposed modules in Pytorch
- block.py has the implementatin of LIST module based DnCNN denoising framework
- Denoising_demo.ipynb is a notebook to reflect our training/inference setup for DnCNN experiments