The improved TEM by more feasible Implementation

In recent years, there are increasing image compression techniques based on machine learning algorithms. Compared to traditional compression mechanism, machine-learning-based image compression techniques are more effective and accurate. However, the TEM algorithm proposed by zhang et al is difficult...

Full description

Saved in:
Bibliographic Details
Published in2017 IEEE 2nd Advanced Information Technology, Electronic and Automation Control Conference (IAEAC) pp. 2545 - 2548
Main Authors Yidan Zhang, Xuedong Liu
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.03.2017
Subjects
active learning
High definition video
image compression
kernel PCA map
point selection
Online AccessGet full text
DOI10.1109/IAEAC.2017.8054483

Cover

More Information
Summary:In recent years, there are increasing image compression techniques based on machine learning algorithms. Compared to traditional compression mechanism, machine-learning-based image compression techniques are more effective and accurate. However, the TEM algorithm proposed by zhang et al is difficult to achieve due to the heavy computation burden. In this paper, we propose an improved algorithm named the improved TEM on the basis of adding seeds interatively for the point selection part to reduce the computational burden. What's more, the kernel PCA map are adopted to diminish the computation complexity through decreasing the dimension of the feature map. The result of feasibility comparison between improved TEM and others shows that improved TEM algorithm is more feasible.
DOI:10.1109/IAEAC.2017.8054483