Abstract
Techniques are presented herein that allow an asynchronous video messaging platform or an online communication and collaboration system to learn, over time, from a user’s behavior to determine when to automatically adjust a video recording’s playback speed and/or volume whenever a speaker (in such a recording) says something that the user would find interesting. Aspects of the presented techniques encompass a gathering of data (regarding, for example, the user’s manipulation of playback speed and volume settings and actions like rewinding or pausing), an analysis of such data (leveraging natural language processing (NLP) techniques to examine video content, such as transcripts or captions, to identify keywords, topics, or sentiments), a modeling of the user’s behavior (including individual preferences, interaction patterns, content preferences, etc.), an offering of specific playback speed and volume setting suggestions, reinforcement learning algorithms that dynamically adjust all of the above, and a collection of operational metrics (regarding accuracy, precision, recall, etc.).
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Gatzke, Alan; Wilson, Andi; and Astashkin, Arsenii, "AI-BASED ADAPTATION OF VIDEO SETTINGS BASED ON LEARNED USER BEHAVIOR", Technical Disclosure Commons, (August 17, 2023)
https://www.tdcommons.org/dpubs_series/6150