Abstract
Intra prediction is a vital part of the image/video coding framework, which is designed to remove spatial redundancy within a picture. Based on a set of predefined linear combinations, traditional intra prediction cannot cope with coding blocks with irregular textures. To tackle this drawback, in this article, we propose a Generative Adversarial Network (GAN)-based intra prediction approach to enhance intra prediction accuracy. Specifically, with the superior non-linear fitting ability, the well-trained generator of GAN acts as a mapping from the adjacent reconstructed signals to the prediction unit, implemented into both encoder and decoder. Simulation results show that for All-Intra configuration, our proposed algorithm achieves, on average, a 1.6% BD-rate cutback for luminance components compared with video coding reference software HM-16.15 and outperforms previous similar works.
Highlights
With the explosive growth of multimedia applications, video traffic accounts for the vast majority of the total network traffic in wired and mobile services [1]
Compared to previous literature specializing in fixed size block: Fully Connected (FC) [15], Convolutional Neural Networks (CNNs) [17] and Recurrent Neural Networks (RNN) [18,19], we proposed a Generative Adversarial Network (GAN)-based intra prediction for a larger block size, which is accompanied by greater prediction difficulty
We propose an intra prediction approach guided by generative adversarial network
Summary
With the explosive growth of multimedia applications, video traffic accounts for the vast majority of the total network traffic in wired and mobile services [1]. It is noteworthy that all the aforementioned intra prediction methods, which use only the closest reconstructed signals for predicting a unit, ignore abundant context between the prediction unit and corresponding adjacent samples and produce inaccurate results especially when weak spatial coherence exists between the prediction unit and the nearest reconstructed signals To tackle this issue, Multiple Reference Line (MRL) [6] intra prediction is adopted in VVC. The neural network-based approaches are implemented into the video compression architecture to strengthen the coding performance of each specific section, such as mode estimation [11,12], partitioning [13,14], intra prediction [15,16,17,18,19,20,21], inter prediction [22] and post-processing [23,24].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
