Subjective video quality assessment for mobile devices
summary
Claim: Adaptive Paired Comparison produces fast, dependable mobile video quality scores by showing each test clip once and using Adaptive Paired Comparison with a particle-filter selection to pick the next reference. Implementation uses a 32-sample test set and a 50-point subjectively linear scale, converges in a few comparisons, yields much smaller confidence intervals than MOS or DS-MOS, runs without a test lab on phones, and produces repeatable numeric quality estimates suitable for regression gating.
AI-written summary · about
chapters (11)
- 0:00Intro
- 0:32Measuring video quality
- 0:55Mean opinion score
- 3:05Double stimulus
- 4:15Pad comparison
- 4:46Linear scale
- 5:56Reference quality
- 6:46Single test
- 8:17Two test results
- 9:05Summary
- 9:39Conclusion
related talks
12mMeasuring live video quality with minimal complexity, now available for everyone!
45mVK Calls: Objectively Assessing Video Quality
48mAudio and Video Testing Fundamentals
27mBehind the Curtains: A/B Tests in Video Land
45mMobile Visual Testing: Uphill Battle Of Mobile Visual Regression
38mDelivering Seamless Mobile Experiences Through Holistic Testing
34mTesting Live Video Streams At Scale
18mTesting and Monitoring the Smartphone Experience