Project 7

• Topic - Image Super Resolution
• Paper - Learning a Single Convolutional Super-Resolution Network for Multiple Degradations
• Paper Link - https://arxiv.org/pdf/1712.06116v2.pdf

Run Instructions

1. make run

Usage

Requirements

• python=3.8
• pytorch=1.7.1
• cudatoolkit=10.2

Code structure

.
├── docs
|   ├── TeamKota-MidEvalReport.pdf
|   ├── Project Proposal
│       ├── Method.pdf
│       └── TeamKota-ProjectProposal.pdf
├── README.md
├── Makefile
├── data
    ├── BSDS300
    ├── DIV2K_train_HR
    ├── waterloo
├── environment.yml
└── src
    ├── dataset.py
    ├── kernels.py
    ├── logger.py
    ├── main.py
    ├── model.py
    ├── train.py
    └── utils.py

Setup env

• conda env create --file environment.yaml

Data

• use Name in main.py

Id	Name	Info	Link
1	DIV2K_train_HR	HR images	http://data.vision.ee.ethz.ch/cvl/DIV2K/DIV2K_train_HR.zip
2	DIV2K_train_LR_bicubic_X2	Bicubic x2 downscaling	http://data.vision.ee.ethz.ch/cvl/DIV2K/DIV2K_train_LR_bicubic_X2.zip
3	Waterloo Exploration Database	Pristine Natural Images	http://ivc.uwaterloo.ca/database/WaterlooExploration/exploration_database_and_code.rar
4	Berkeley Segementation Dataset and Benchmark (BSDS300)	Natural Images with greyscale and color segementation	https://www2.eecs.berkeley.edu/Research/Projects/CS/vision/bsds/BSDS300-images.tgz