For a wide range of companies, delivering caliber rich media experiences is of paramount grandness. However, mensurate the quality of video that witness really experience has been difficult, if not impossible, in many circumstances where it is not possible to explicitly ask spectator to rate the quality of the video, or when a acknowledgment video is not available.

HeadSpin ’ s patent-pending reference-free video MOS or Mean Opinion Score provides a flexible, exact, and scalable alternative to traditional survey-based and full-reference algorithmic attack, using to address the challenges of reliably assessing perceptual video quality.

In eminent quality wildlife video, the blurriness of the background is a deliberate result of focusing on the animal. This is an example of something an AI deep erudition model could be trained to recognize but would be lose entirely by a parametric model based on video character metrics.

Importance of Video Quality

Poor experience matter. They lead to dissatisfied users, and dissatisfied users don ’ t stay dissatisfied for long—they go elsewhere. AnAkamai accountlaunch that a individual buffering event induce user happiness to drop by 14 %. Conversely, higher bit rate can boost viewer engagement by more than 10 %.

Today, rich media content, including live picture streams on social platform, picture conference and chat service, nomadic gaming, and television broadcasts, represents the lion ’ s share of network traffic and a major part of the user experience. Poor digital experience could be caused by an issue with the picture rootage, network conditions that get packet loss and streaming hold, and even device-specific issues that create provide job. When experiences are suboptimal, any number of element across a orbit of domains could be the perpetrator and it ’ s vital for administration to be able to understand—and optimize—the caliber of picture content delivery on real device in various locations, in real clip.

Traditional Video Quality Monitoring Approaches

Video quality monitoring has traditionally be assessed using a metric cognize as the —a subjective amount of perceptual video calibre monitoring as grade by a jury of users. MOS gobs are usually gathered in a caliber evaluation test, such as when WhatsApp inquire you to rate the quality of your video call, but they can also be algorithmically predicted in the absence of real user feedback.

For scenarios in which a credit video is available, there are existing standards that can be used to establish a video quality monitoring MOS:

• International Telecommunication Union Radiocommunication Sector (ITU-R).The ITU-R has well-defined standards for developing experiments to estimate video quality MOS. Using the measure, video character MOS can be compute based on argument deduce from both the reference video and the test video. An idea of the video MOS is calculated from a parametric model constructed using these picture metadata parameters regressed against immanent quality scores from user survey.
• Video Multimethod Assessment Fusion (VMAF).VMAF is an open beginning algorithm developed by Netflix that uses a strict frame-by-frame equivalence of reference and observed picture in order to make a prediction of the picture quality MOS. This approach, while offering many benefits, is only meaningful across variants of a single reference picture, e.g., multiple squeeze versions of a video.

With both of these standards, teams are face with significant limitation.

1. There are many scenarios in which it simply isn ’ t possible to establish a reference video.This is true in scenarios, live video streaming, video calls, and a number of other cases in which content is dynamically generated.
2. Reference picture are often too costly or resource-intensive to create and to preservestill if they may exist. This is often true in broadcast television and many video streaming scenarios.
3. MOS grade from full-reference approaches can not be meaningfully equate across different source files. They are typically exclusively useful for a limited act of use cases, such as video compression optimization, when there is a need to compare across variants of a single citation video.
4. Existent macrocosm use example often break fundamental premiss in full-reference video MOS techniques. For example, even if the beginning picture is the like, the blind transcription of the picture playback on the device is go to be slightly different for each playback. This was part of the motive for our reference-free MOS.
5. Many full-reference MOS scores are not very easily anchor in the perceived quality. For instance, if the reference video itself is low quality, VMAF will produce a high score where a real user would not.
6. Video technology and the consumer-facing content environ that technology are forever evolving. The existing standards and result described supra experience no strategy for continuous improvement and evolution as technology standards continue to vary.

These limitations pose significant challenges as team continue to focus on render high quality rich media substance.

Introducing the HeadSpin Reference-Free Video MOS

To direct the challenges and limitations of traditional approaching to MOS, HeadSpin ’ s data science team develop a amply AI-based reference-free video MOS framework leveraging experience from the 5+ geezerhood we ’ ve been in business. Our innovative patent-pending algorithm offers breakthroughs in the measurement of video lineament by providing a reference-free estimation of the true subjective lineament score of video content as perceived by end-users based on.

For autonomous testing across multiple user personas, check out SUSATest — it explores your app like 10 different real users.

The reference-free approach enables test/QA and maturation teams to quickly and cost-effectively trial, monitor, and analyze, at scale, the lineament of video being deliver.

Key Features

Simple, Scalable Implementation for Better Video Quality Monitoring
HeadSpin ’ s MOS may be applied to any video content regardless of origin. It act seamlessly across devices, on picture captured directly on the program, and still on third party videos imported to the platform through our API. A MOS score is guess for every frame in the supplied video based on spatial and temporal features extracted from the video.

Video Quality Monitoring With Reference Free Analysis
Our reference-free approach enable our users to judge video quality without any reference rootage video or comparison processes. In addition, the HeadSpin program also offers a comprehensive suite of reference-free video lineament metric that track multiple picture features such as blockiness, blurriness, and line. These extra metrics are stage in a time-aligned view alongside the picture quality MOS time series and can be used to diagnose or explain regions that exhibit poor video quality scores. When geminate with our expert scheme AI analysis, such as thePoor Video Quality Issue Card, the resolution will surface insights into perceptual video quality issues, and into correlativity between these issues and other app-related metrics.

HeadSpin surfaces insights into perceptual video quality issues in a time-aligned view—highlighting correlations between frame-by-frame MOS scores and other video- and app-related metrics.

Improved Video Quality Monitoring With A Flexible, Accurate ML Model

‍Our reference-free video MOS is non-parametric and employs convolutional neural network technology to expose perceptual video character features. It is the solitary MOS on the market that perform not explicitly rely on other prosody for results. Unlike a parametric approach derived from video quality metrics (such as blurriness, blockiness, jerkiness, etc.), which is prone to mistaken positives, for example, from substantial or orthogonal UI constituent, splash screens, logo, stylistic pixelation, semi-transparent elements, and gage special outcome, our AI-based MOS has be educate to spot these scenarios. In add-on, the AI will accurately place a picture stream that is potential to be perceived as low-quality even if it is captured at high resolution. This unique approach gives HeadSpin the exemption and flexibility to capture and identify video lineament issues that are not currently rise by existing standards like ITU-R and VMAF.

‍Unprecedented Wealth of Data for Accurate Video Quality Monitoring
HeadSpin has conduct an innovative approach to leverage our expertness and resources to make a model anchor in real world picture pullulate use cases so that the character of the picture can be estimated without a reference compare. We curated the largest video quality information set of its kind consisting of M of unique videos get under real-world weather via the HeadSpin. The AI model was trained on a subset of the videos beguile and we have & gt; 60,000 labels on & gt; 700 videos from user study. In price of diverseness of unique videos, various content from a diverse set of content supplier, and sourcing labels from high caliber labelers, our reference-free video MOS model is establish on the most comprehensive dataset in the infinite.

Constant Improvement Toward Video Quality Monitoring With Continuous Optimization
Due to the nature of our deep learning architecture, our models continue to evolve. Incorporating data from the latest picture teem covering enables the machine con framework to accurately predict MOS on any picture it canvas. Continuously incorporate customer feedback into the model development operation (via our Video Annotation App) enable us to ameliorate the accuracy of the AI poser over time.

Better Together
HeadSpin additionally indorse a full-reference VMAF MOS on the platform. While VMAF is more suited for bespeak degradation relative to the source, our reference-free MOS will indicate absolute video quality as perceived by the end user. Many of our users use both in tandem as complementary measures.

Video MOS Use Cases

Below are a few instance of the many ways our customers are employing HeadSpin ’ s video quality MOS.

Measuring Live Video Streaming
Live video streaming substance is growing in popularity every day. More and more platforms are offering exploiter the power to both serve and waste alive picture message. Tracking the perceived character of these live streaming videos is significant to understanding how users perceive the calibre of the program itself. With its reference-free approach, the HeadSpin video MOS helps our users translate the video calibre of alive streaming content.

Reference-based Video Streaming Measurement
In cases where reference picture are available, HeadSpin ’ s deterministic AI algorithms can be use to compare a subsequent or parallel test against the reference to supply a comparison of the video quality. Users frequently set up taxonomical tests across devices, locations, and carriers, on repeated intervals, to get a best agreement of how these variable affect the percept of the video substance being served from their platform.

Statistical Analysis
Our users frequently leverage our reference-free video MOS algorithm to combine the MOS over many experiments. This aggregated intelligence can be used to develop statistical methods for testing hypotheses. For example, they can use our AI platform to quiz a hypothesis that streaming video on a mobile app has greater division in character during peak internet usage than during non-peak multiplication. HeadSpin can be expend to instrument this experiment and collect the MOS information required to test this hypothesis.

To calibrate the accuracy of the human-annotated values used in our ML model, a tight exercise was direct with a jury of information expert not imply in the annotation study, to come to a consensus on “ ground-truth ” MOS scores.

Conclusion

It is not uncommon for picture content to receive 1000000 of views every day. How these videos are perceived by the end user may rely on a multitude of factors, including content delivery network (CDN) configurations, picture encoding and playback optimizations, network conditions, and more. In order to render best picture hosting and serving services, it is lively to understand how different program settings may touch the quality of video viewers see, and to understand how that experience may deviate across devices and locations.

For many organizations today, picture quality is too crucial to leave to fortune or guesswork. With HeadSpin ’ s reference-free video MOS, app teams can derive the accuracy and scalability they need to track and optimise the lineament of the witness ’ experience.

The provides execution and quality-of-experience analytics for mobile, web, audio, and video. With HeadSpin, technology, QA, operations, and ware team can assure optimal digital experiences throughout the ontogeny lifecycle.

Start measure your video quality with reference-free simplicity today! Speak to a salesperson toget started.

FAQs

1. What can AI accomplish with video apps?

AI-based can recommend content or synergistic stories. Data mining on end-user behavior enables game architect to enquire how players use the game, what areas they play the most, and what lead them to stop playing, let the developer to modify gameplay or enhance monetisation.

2. What component are considered while measuring picture quality?

Factors that are considered to analyse video quality are:

• Bit Rate
• Buffer filling
• Lag duration
• Play length
• Lag ratio

3. What element are measured in video load testing?

With the help of an AI-backed video testing platform, it is possible to perform load testing for video apps and monitor:

• crucial video streaming prosody, include adaptive bitrate cyclosis
• real-time viewership and preparing for peak viewership scenarios
• viewing behavior of users and consequently develop video cyclosis infrastructure

4. What are the HeadSpin AV platform features that help in testing video apps?

HeadSpin ’ s AI-powered AV solution has robust features such as:

• Cross-device and browser compatibility- for running exam on your OTT and other medium applications like gaming, picture conferences, and live streaming platform across a all-inclusive compass of devices.
• QoE & amp; Streaming Performance KPIs- a comprehensive suite of reference-free quality prosody to examine and assure blockiness, blurriness, luminance, colorfulness, exposure, flickering, freeze, video frame pace drib, and buffering clip
• Perceptual video quality analysis- based on predictive machine learning model to deliver exact immanent analysis with a time aligned-view of video lineament as perceive by end-users.

4 Parts

Connet Now