Optimization of Remote Desktop with CNN-based Image Compression Model

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

Remote desktop systems become commonly used for users to enhance the efficiency of their daily tasks commonly. In this work, we propose an expanded image compression model with convolutional neural network (CNN) and train two jointly optimized CNN based models as the image encoder and decoder to optimize the compression of the desktop images and design a new compartmentalization of the update desktop region to fit the CNN encoder. We implement the proposed encoding on the open source Remote Frame Buffer (RFB) protocol. Compared with tight encoding which is dedicated to low-bandwidth remote desktop, the proposed encoding method prompts the user experience with a even lower network bandwidth consumption.

Original languageEnglish
Title of host publicationKnowledge Science, Engineering and Management - 14th International Conference, KSEM 2021, Proceedings
EditorsHan Qiu, Cheng Zhang, Zongming Fei, Meikang Qiu, Sun-Yuan Kung
PublisherSpringer Science and Business Media Deutschland GmbH
Pages692-703
Number of pages12
ISBN (Print)9783030821357
DOIs
StatePublished - 2021
Externally publishedYes
Event14th International Conference on Knowledge Science, Engineering and Management, KSEM 2021 - Tokyo, Japan
Duration: 14 Aug 202116 Aug 2021

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12815 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference14th International Conference on Knowledge Science, Engineering and Management, KSEM 2021
Country/TerritoryJapan
CityTokyo
Period14/08/2116/08/21

Keywords

  • Convolutional neural network
  • Remote desktop
  • Remote frame buffer
  • VNC

Fingerprint

Dive into the research topics of 'Optimization of Remote Desktop with CNN-based Image Compression Model'. Together they form a unique fingerprint.

Cite this