Abstract
Video services are meant to be a fundamental tool in the development of oceanic research. The current technology for underwater networks (UWNs) imposes strong constraints in the transmission capacity since only a severely limited bitrate is available. However, previous studies have shown that the quality of experience (QoE) is enough for ocean scientists to consider the service useful, although the perceived quality can change significantly for small ranges of variation of video parameters. In this context, objective video quality assessment (VQA) methods become essential in network planning and real time quality adaptation fields. This paper presents two specialized models for objective VQA, designed to match the special requirements of UWNs. The models are built upon machine learning techniques and trained with actual user data gathered from subjective tests. Our performance analysis shows how both of them can successfully estimate quality as a mean opinion score (MOS) value and, for the second model, even compute a distribution function for user scores.
Highlights
Video quality assessment (VQA) is an important aspect of multimedia services in any communication network
In [2] the authors have already shown that video services are feasible, even with the high constraints of the current technology, but within a small range of the considered input variables the quality jumps from the bottom to the top of the quality scale
This agrees with conclusions presented in [3] where the authors state that models for objective quality assessment achieve a better performance when they are tuned for a combination of human and system factors
Summary
Video quality assessment (VQA) is an important aspect of multimedia services in any communication network. Objective methods yield the same quality estimation every time a certain video sample is given as input. In [2] the authors have already shown that video services are feasible, even with the high constraints of the current technology, but within a small range of the considered input variables the quality jumps from the bottom to the top of the quality scale This agrees with conclusions presented in [3] where the authors state that models for objective quality assessment achieve a better performance when they are tuned for a combination of human and system factors. Since we have available a dataset with subjective quality scores for underwater video [2], we can leverage the power of the machine learning approach to build regression models for objective quality estimation.
Sign-in/Register to view highlights & summary 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 callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
