ABSTRACT
Contents 13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372 13.2 Performance Evaluation of Quality Assessment Algorithms . . . . . . . . . . . . 374 13.3 Display-Specific Quality Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
13.3.1 Large-Screen Displays. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379 13.3.2 Handheld Displays. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383 13.3.3 3D Displays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
13.4 Content-Specific Quality Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 13.5 Current and Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 13.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
13.1 Introduction According to a recent study, the amount of video traffic streamed over the Internet accounts for 51% of the total amount of data streamed over the Internet as against ∼10% 5 years ago [5]. A large number of portable devices capable of streaming and playing videos over wireless/3G/4G networks are being produced and consumed at a rapidly increasing pace. It is obvious that in today’s world, visual information, especially that transmitted over the Internet, is a major contributor to network traffic. As recent caps on bandwidth announced by AT&T demonstrate, current network capacities are not capable of managing the amount of traffic that such streaming generates. One possible solution to this problem of increasing bandwidth demand is to evaluate whether the current bandwidth allocated to a video stream is necessary for maintaining a certain level of perceptual video quality at the receiver. Specifically, the question that one seeks to answer is, if one reduces the bandwidth for a video stream by a factor of x, how does this affect perceptual video quality? Related to this is another relevant question-by what factor can one reduce the bandwidth of a video stream without any change in perceptual quality?
