1 / 15

Time-Compression: System Concerns, Usage, and Benefits

Time-Compression: System Concerns, Usage, and Benefits. Nosa Omoigui, Li-wei He, Anoop Gupta, Jonathan Grudin, and Elizabeth Sanocki Microsoft Research. Introduction. Digital multimedia content is becoming widespread Unlike text, multimedia is not easy to skim

lynley
Download Presentation

Time-Compression: System Concerns, Usage, and Benefits

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Time-Compression:System Concerns, Usage, and Benefits Nosa Omoigui, Li-wei He, Anoop Gupta, Jonathan Grudin, and Elizabeth Sanocki Microsoft Research

  2. Introduction • Digital multimedia content is becoming widespread • Unlike text, multimedia is not easy to skim • Ability to skim is a key advantage of digital multimedia • Time-compression • Skimming • We focus on time-compression

  3. Time-Compression Basics • Reduces the time it takes to listen to or watch multimedia • Audio time-compression • Signal processing issues are well understood • Re-sampling results in an increase in pitch • Pitch-preserving techniques: • Segment abutting, OLA, SOLA, P-SOLA • Video time-compression • Relatively straightforward • Drop frames or change frame rendering rate

  4. Demo

  5. Time-Compression Issues • Signal processing technology already exists • However, not found in client-server environments • Several possible design techniques • Multiple server-side files • Simple client-side solution • Sophisticated client-side solution • Systems concerns include speedup-factor granularity, latency, storage, and bandwidth

  6. Multiple Server-Side Files • Advantages • No additional bandwidth demands • Relatively simple to implement • Disadvantages • Discrete speed-up factor granularity • Latency while switching speedup-factors • Requires additional storage • Does not work with existing media files

  7. Simple Client-Side Solution • Advantages • Continuous granularity • No additional storage demands • Works with existing media files • Disadvantages • Requires additional bandwidth • Some client-side complexity • Latency while switching speedup-factors

  8. Sophisticated Client-Side Solution • Advantages • Continuous granularity • No additional storage demands • Works with existing media files • No latency while switching speedup-factors • Disadvantages • Requires additional bandwidth • Significant client-side complexity

  9. Focus of Our Study • Impact of engineering trade-offs on user behavior • Speedup factor adjustment granularity • Latency for speedup adjustment • User behavior: time saved, number of adjustments, perception on the feature, etc.

  10. Experimental Procedure • Dimensions for comparison • Continuous vs. discrete speedup factor adjustment • Long latency vs. short latency • User study of 15 subjects • Task: watch video and give a verbal summary • Five 25-40 minute videos • 2 Discovery channel videos and 3 presentations from ACM 97 • Five conditions (within subject; counter-balanced)

  11. Speedup Factor Achieved • Results • No significant differences across conditions • Average speedup factor is 1.42 • Savings of about 45 minutes for 2 hours of video • Substantial variation across subjects (1.18 – 1.65) • Key implication • Implementers should feel free to choose the simplest solution

  12. Savings in Task Time Are Significant • Our question • Would time for reviewing outweigh time saved using compression? • Method • Instrument the UI to log subjects’ action • Divide the task time into 5 categories: • View, review, pause, seek, and latency • Results • Review time goes up with time-compression • 9-11% vs. 6% • Still substantial savings in task time achieved • 22% savings over no-time-compression

  13. Number of Speedup Adjustments • Our hypotheses • More adjustments at the beginning of the talk • More adjustments for the lower-latency conditions • Observations • Adjustments do tend to occur more in the beginning • No significant difference across conditions

  14. User Feedback and Comments • 87% either loved the feature or found it very useful • Experienced subjects found the normal speed too slow • “Once people have experienced time compression, they will never want to go back” • Employ time-compression with other features • E.g. bookmarks, table of content, and instant-replay button

  15. Conclusions • Subjects loved the time-compression feature • Significant savings in task time is achievable • Our study shows 22% savings • Implementers should feel free to choose the simplest solution of the three trade-offs • No significant differences across conditions

More Related