Video Quality Prediction Under Time-Varying Loads

Izima, Obinna; Fréin, Ruairí de; Davis, Mark

doi:10.1109/cloudcom2018.2018.00035

Cited by 4 publications

(7 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We also note that the survey missed out on some novel ML applications in networking. For instance, the load-adjusted learning first proposed in [34], and used in [9,40,63], was not covered in the survey. Our background in algorithms as opposed to networks gives us a different perspective on the survey.…”

Section: Overview Of ML Applications In Networkmentioning

confidence: 99%

“…Figure 5 depicts a time-varying client population accessing video streams being streamed by a cloud-hosted video server. Using this setup, the authors of [9,63] demonstrated how audio buffer rate (ABR), video frame rate (VFR), and packet count, QoD measurements, could be predicted using system activity reports (SAR). The SAR function is used to extract device statistics from the server.…”

Section: Overview Of ML Applications In Networkmentioning

confidence: 99%

“…The EN and RF models achieved the best performances for all QoD targets. Figure 6, R2 compares the prediction accuracy of the EN model with the actual packet count statistics [9]. By utilizing a suitable adaptive regularization function, the load-adjusted learning algorithms outperformed the techniques that did not model the load effect by 10%, and reduced computation.…”

Section: Overview Of ML Applications In Networkmentioning

confidence: 99%

“…The authors reported an application-aware QoD prediction accuracy of over 90%. However, given that the setup for this was similar to [9], as was the feature set, incorporating the system load [40] would have better modeled the problem, as demonstrated by the work in [9]. The study in [179] utilized regression ML models to predict mean opinion score (MOS).…”

Section: Software-defined Networking (Sdn)mentioning

confidence: 99%

“…ML delivers insights to a wide range of disciplines, including computer vision [3], speech recognition [4,5], software defined networks (SDNs) [6], communication networks [7], and Internet of Things (IoT) [8]. In recent years, there has been a rapid adoption of ML techniques to solve practical problems in computer networks [9]. In this paper, we provide a review of recent applications of ML techniques that use quality of delivery (QoD) measurements for the prediction of video quality.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A Survey of Machine Learning Techniques for Video Quality Prediction from Quality of Delivery Metrics

2021

Self Cite

View full text Add to dashboard Cite

A growing number of video streaming networks are incorporating machine learning (ML) applications. The growth of video streaming services places enormous pressure on network and video content providers who need to proactively maintain high levels of video quality. ML has been applied to predict the quality of video streams. Quality of delivery (QoD) measurements, which capture the end-to-end performances of network services, have been leveraged in video quality prediction. The drive for end-to-end encryption, for privacy and digital rights management, has brought about a lack of visibility for operators who desire insights from video quality metrics. In response, numerous solutions have been proposed to tackle the challenge of video quality prediction from QoD-derived metrics. This survey provides a review of studies that focus on ML techniques for predicting the QoD metrics in video streaming services. In the context of video quality measurements, we focus on QoD metrics, which are not tied to a particular type of video streaming service. Unlike previous reviews in the area, this contribution considers papers published between 2016 and 2021. Approaches for predicting QoD for video are grouped under the following headings: (1) video quality prediction under QoD impairments, (2) prediction of video quality from encrypted video streaming traffic, (3) predicting the video quality in HAS applications, (4) predicting the video quality in SDN applications, (5) predicting the video quality in wireless settings, and (6) predicting the video quality in WebRTC applications. Throughout the survey, some research challenges and directions in this area are discussed, including (1) machine learning over deep learning; (2) adaptive deep learning for improved video delivery; (3) computational cost and interpretability; (4) self-healing networks and failure recovery. The survey findings reveal that traditional ML algorithms are the most widely adopted models for solving video quality prediction problems. This family of algorithms has a lot of potential because they are well understood, easy to deploy, and have lower computational requirements than deep learning techniques.

show abstract

Section: Overview Of ML Applications In Networkmentioning

confidence: 99%

Section: Overview Of ML Applications In Networkmentioning

confidence: 99%

Section: Overview Of ML Applications In Networkmentioning

confidence: 99%