Abstract
Speeding up video playback without taking into account the semantic context can result in degradation of audio and motion quality. This disclosure describes techniques that optimize audio and video playback by semantically analyzing content to selectively skip irrelevant segments. Audio and/or video content is segmented into temporal chunks, classified based on contextual importance using machine learning models, and adjusted according to a target playback speed. Content is smoothed for continuity and sped up at lower compression factors to maintain quality. Optionally, skipped but contextually significant segments are replaced with annotations or visual cues to preserve semantic coherence. The described techniques combine semantic filtering/ content summarization with traditional playback speed-up methods, enabling efficient content consumption while maintaining user experience and content integrity.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Tran, Duc-Hieu and Hartmann, Florian, "Semantic Speed-up of Audio and Video", Technical Disclosure Commons, (June 17, 2025)
https://www.tdcommons.org/dpubs_series/8242