Abstract
Video content can include portions that are inappropriate for certain groups of viewers, e.g., children. While parental controls are available, these can be ineffective for online video content from video creators and do not provide personalization to individual viewers. This disclosure describes the use of artificial intelligence techniques to automatically analyze portions of a video that are coming up for playback to generate a summary. The summary output by the model is evaluated against pre-configured guardrails provided by a parent. The pre-configured guardrails can be specified via natural language input. If the summary matches the guardrails, indicating that the upcoming portion of a video likely depicts content of a type that the parent has restricted from access, the portion is suppressed (e.g., not shown/skipped, masked) by the media playback software/hardware.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Raman, Sundarrajan, "Natural Language Based Video Guardrails for Parental Control", Technical Disclosure Commons, (August 09, 2024)
https://www.tdcommons.org/dpubs_series/7270