Image Super-Resolution with Deep Convolutional Neural Network
Authors:
- Xiancai JiXJXiancai Ji
- Yao LuYLYao Lu
- Li GuoLGLi Guo
- 1 June 2016
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 626-630
Abstract
We present a computational model for image super-resolution. Apart from using deep Convolutional Neural Network to map between the low-resolution images and high-resolution images, we adopt stepwise refinement method to improve the reconstruction results and introduce the back projection algorithm to avoid diffusion of error pixels.Keywords
This publication has 10 references indexed in Scilit:
- Spatial Pyramid Pooling in Deep Convolutional Networks for Visual RecognitionIEEE Transactions on Pattern Analysis and Machine Intelligence, 2015
- Caffe2014
- Visualizing and Understanding Convolutional Networks2014
- Single-Image Super-Resolution: A Benchmark2014
- Learning a Deep Convolutional Network for Image Super-Resolution2014
- Anchored Neighborhood Regression for Fast Example-Based Super-Resolution2013
- Fast Image Super-Resolution Based on In-Place Example Regression2013
- Image denoising: Can plain neural networks compete with BM3D?2012
- Image Super-Resolution Via Sparse RepresentationIEEE Transactions on Image Processing, 2010
- Example-based super-resolutionIEEE Computer Graphics and Applications, 2002