Wednesday, November 30, 2011
What:
Dialogic proposes an alternative, patent pending approach to predicting video quality by presenting the paper, Support Vector Regression-Based Video Quality Prediction.
Why:
Traditional methods for estimating video quality include models that simulate the Human Visual System (HVS), which enables the human eye and brain to process visual details. Unfortunately, not all HVS simulation models have been effective enough to yield valuable results, and those that are reliable have tended to be cost-prohibitive. This has led engineers in some cases – particularly where resources are limited - to look instead to methods such as G.1070-based video quality prediction, which can lead to inaccuracy.
Given this situation, the industry needs a method of estimating video quality that requires fewer computational resources, but still provides accurate approximations of video quality. A Support Vector Machine (SVM) supervised learning approach directly addresses this need. Instead of using extracted features as input to a model (as with HVS models), Dialogic proposes using the extracted features to build a quality measure with SVM supervised learning. In addition to being cost-effective, this patent pending approach can better approximate the National Telecommunications and Information Association (NTIA) Video Quality Metric (VQM) and subjective Mean Opinion Score (MOS) values as compared to G.1070-based video quality prediction.
Who:
Dekun Zou, Senior Scientist at Dialogic
When:
Monday, December 5, 2011
10:00 - 11:40 a.m. Pacific Time
Where:
Laguna Cliffs Marriott Resort and Spa
25135 Park Lantern
Dana Point, CA 92629
About Dialogic
For more than 25 years, Dialogic (NASDAQ: DLGC) and its subsidiaries have been providing communications platforms and technology to enterprise and service provider markets. Our portfolio of IP and TDM based multimedia processing and call control technologies enables developers and service providers to build and deploy innovative applications without concern for the complexities of the communications medium or network. This empowers our customers to unleash the profit from video, voice and data for advanced networks.
For more information on Dialogic, visit www.dialogic.com. Also, find us on the following social networking sites:
- Dialogic Exchange Network (DEN)
- Facebook
- Twitter
- YouTube
Dialogic is a registered trademark of Dialogic Inc. or a subsidiary thereof (“Dialogic”). Other trademarks mentioned and/or marked herein belong to their respective owners. Dialogic and its subsidiaries encourage all users of their products to procure all necessary intellectual property licenses required to implement their concepts or applications, which licenses may vary from country to country.