Image super-resolution via a novel cascaded convolutional neural network framework
作者：
Highlights：
•
• Novel cascaded CNN framework is designed for the multi-scale image SR task with a single trained model.
• Multi-scale feature mapping is proposed to extract the inherent features via the low-resolution image.
• Parallel network architecture is designed to predict more feature maps.
• Comparison results indicate that the proposed framework achieves good image SR performance.
摘要
•Novel cascaded CNN framework is designed for the multi-scale image SR task with a single trained model.•Multi-scale feature mapping is proposed to extract the inherent features via the low-resolution image.•Parallel network architecture is designed to predict more feature maps.•Comparison results indicate that the proposed framework achieves good image SR performance.
论文关键词：Image super-resolution,Cascaded convolution neural network,Multi-scale feature mapping,Residual learning,Gradient clipping
论文评审过程：Received 30 July 2017, Revised 18 December 2017, Accepted 27 January 2018, Available online 3 February 2018, Version of Record 6 February 2018.
论文官网地址：https://doi.org/10.1016/j.image.2018.01.009
