Published on Sat Apr 18 2020

ImagePairs: Realistic Super Resolution Dataset via Beam Splitter Camera Rig

Hamid Reza Vaezi Joze, Ilya Zharkov, Karlton Powell, Carl Ringler, Luming Liang, Andy Roulston, Moshe Lutz, Vivek Pradeep

Super Resolution is the problem of recovering a high-resolution image from a single or multiple low-resolution images of the same scene. We are proposing a new data acquisition technique for gathering real image data set which could be used as an input for super resolution, noise cancellation and quality enhancement techniques.

0
0
0
Abstract

Super Resolution is the problem of recovering a high-resolution image from a single or multiple low-resolution images of the same scene. It is an ill-posed problem since high frequency visual details of the scene are completely lost in low-resolution images. To overcome this, many machine learning approaches have been proposed aiming at training a model to recover the lost details in the new scenes. Such approaches include the recent successful effort in utilizing deep learning techniques to solve super resolution problem. As proven, data itself plays a significant role in the machine learning process especially deep learning approaches which are data hungry. Therefore, to solve the problem, the process of gathering data and its formation could be equally as vital as the machine learning technique used. Herein, we are proposing a new data acquisition technique for gathering real image data set which could be used as an input for super resolution, noise cancellation and quality enhancement techniques. We use a beam-splitter to capture the same scene by a low resolution camera and a high resolution camera. Since we also release the raw images, this large-scale dataset could be used for other tasks such as ISP generation. Unlike current small-scale dataset used for these tasks, our proposed dataset includes 11,421 pairs of low-resolution high-resolution images of diverse scenes. To our knowledge this is the most complete dataset for super resolution, ISP and image quality enhancement. The benchmarking result shows how the new dataset can be successfully used to significantly improve the quality of real-world image super resolution.

Thu Aug 05 2021
Computer Vision
Dual-reference Training Data Acquisition and CNN Construction for Image Super-Resolution
Low and high resolution image pairs synthesized by existing degradation models deviate from those in reality. The super-resolution CNN trained by these synthesized LRHR image pairs does not perform well when being applied to real images.
11
0
0
Wed Mar 03 2021
Computer Vision
Real-World Single Image Super-Resolution: A Brief Review
Single image super-resolution (SISR) aims to reconstruct a high-resolution image from a low-resolution observation. This article aims to make a comprehensive review on RSISR methods. It covers the critical publically available datasets and assessment metrics.
0
0
0
Mon Apr 01 2019
Computer Vision
Toward Real-World Single Image Super-Resolution: A New Benchmark and A New Model
Most of the existing learning-based single image superresolution (SISR) methods are trained and evaluated on simulated datasets. We build areal-world super-resolution (RealSR) dataset where paired LR-HR images on the same scene are captured by adjusting the focal length of a digital camera.
0
0
0
Sun Jun 21 2020
Machine Learning
Mapping Low-Resolution Images To Multiple High-Resolution Images Using Non-Adversarial Mapping
The Single Image Super-Resolution (SISR) problem is an one-to-many mapping problem. To circumvent this problem, we propose SR-NAM which utilizes the Non-Adversarial Mapping (NAM) technique.
0
0
0
Tue Feb 09 2021
Computer Vision
Deep learning architectural designs for super-resolution of noisy images
State-of-the-art methods often fail atconstructing high-resolution images from noisy versions of their low-resolution counterparts. In this work, we propose to jointly perform denoising and super-resolution.
0
0
0
Fri Jul 02 2021
Computer Vision
Unsupervised Single Image Super-resolution Under Complex Noise
The real image degradation is usually unknown and highly variant from one to another. This makes it extremely hard to train a single model to handle the general SISR task. Instead of the traditional i.i.d. Gaussian noise distribution, a novel patch-based non-i.D. noise modeling method is proposed to fit the complex real noise.
4
0
0