Project

A Study on Optimal Design of Multi-view Coded Image Quality Assessment Based on Fast Super-Resolution Convolutional Neural Network

Goal: To propose optimal design of coded image quality assessment of multi-view and super-resolution images based on deep learning (Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), etc...).

Methods: Superresolution, Image Quality Assessment, Multi-view 3D Image, Convolutional Neural Network, FSRCNN

Date: 1 April 2018

Updates
0 new
3
Recommendations
0 new
0
Followers
0 new
1
Reads
0 new
9

Project log

Norifumi Kawabata
added a research item
We can come to approach on super-resolution processing based on deep learning by appearing deep learning tools. These performance are shown by applying the only deep learning theory for super-resolution processing. However, we consider that the optimal condition and design for super-resolution processing are achieved better by improving these algorithms and setting parameter appropriately. In this paper, first, we carried out experiments on optimal condition and design of super-resolution processing for the multi-view 3D images encoded and decoded by H.265/HEVC, focused on structure of convolutional neural network by using Chainer. And then, we assessed for the generated images quality objectively, and compare to each image. Finally, we discussed for experimental results.
Norifumi Kawabata
added an update
One international conference paper has been accepted to the 2018 7th IEEE Global Conference on Consumer Electronics (GCCE2018) !
Norifumi Kawabata, “HEVC Image Quality Assessment of the Multi-view and Super-resolution Images Based on CNN," Proc. of 2018 IEEE 7th Global Conference on Consumer Electronics (GCCE 2018), 2 pages, Nara Royal Hotel, Nara, Japan, October 9-12, 2018 (Accepted).
 
Norifumi Kawabata
added a research item
We can come to approach on super-resolution processing based on deep learning by appearing deep learning tools. These performance are shown by applying the only deep learning theory for super-resolution processing. However, we consider that the optimal condition and design for super-resolution processing are achieved better by improving these algorithms and setting parameter appropriately. In this paper, first, we carried out experiments on optimal condition and design of super-resolution processing for the multi-view 3D images encoded and decoded by H.265/HEVC, focused on structure of convolutional neural network by using Chainer. And then, we assessed for the generated images quality objectively, and compare to each image. Finally, we discussed for experimental results.
Norifumi Kawabata
added an update
I will have been participated and presented in Image Media Quality Technical Meeting held at Tsudanuma Campus, Chiba Institute of Technology on May 25, 2018.
 
Norifumi Kawabata
added a project goal
To propose optimal design of coded image quality assessment of multi-view and super-resolution images based on deep learning (Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), etc...).