Video Relevance Identification via Transcript Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Users often play irrelevant portions of long videos while searching for relevant content, leading to excessive consumption of electronic resources such as network bandwidth and device battery.

Innovation Solution

A system that identifies relevant portions of a video by generating a transcript using a speech-to-text algorithm, allowing users to select and share specific points of interest, and displays user-approved text alongside the video, enabling efficient playback by starting from the relevant point.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If users play through long videos to find relevant content, then users can access relevant information, but electronic resources (network bandwidth, battery) are excessively consumed

Engineering Contradiction:
Improverelevant content accessibilityVSAvoidelectronic resource consumption
Core Design Contradiction:
Loss of informationVSLoss of energy

Solution Approach 1:

The patent segments the video content by generating a transcript that divides the video into distinct textual portions corresponding to different time segments. Users can search and navigate to specific segments of interest without playing through the entire video, thereby reducing energy consumption while maintaining access to relevant content.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a transcript as an intermediary between the user and the video content. The transcript serves as a searchable text representation that allows users to identify relevant portions without directly playing the video, reducing unnecessary video playback and associated resource consumption.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Loss of information

If users play irrelevant portions of video, then users can find relevant content, but time is wasted on unnecessary playback

Engineering Contradiction:
Improverelevant content discoveryVSAvoidvideo playback time
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The patent performs preliminary action by generating a complete transcript of the video before the user needs to search for content. This pre-processed text representation enables users to quickly search and identify relevant sections without having to play through irrelevant portions, saving time while maintaining content discoverability.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent replaces the mechanical video playback system with a text-based search system. Instead of playing video content sequentially and hoping to find relevant portions, users can search the transcript text directly, substituting the time-consuming video playback mechanism with a more efficient text search approach.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS11463748B2Identifying relevance of a video
Publication Date: 2022.10.04 MICROSOFT TECHNOLOGY LICENSING LLC
  • US11463748B2 patent drawing
  • US11463748B2 patent drawing
  • US11463748B2 patent drawing

AI summary

Techniques for identifying relevance of a video are disclosed herein. In some embodiments, a computer-implemented method comprises: causing a video to be played on a device of a user; receiving, from the device, an instruction to share the video with another user, the instruction corresponding to a point-in-time in the video; identifying text in a transcript of the video based on the point-in-time; causing the identified text to be displayed on the device based on the instruction to share the video; receiving, from the device, an instruction to include user-approved text along with the video in the sharing of the video with the one or more other users, the user-approved text comprising at least a portion of the identified text; and causing the user-approved text to be displayed on a device of the other user in association with the video based on the instructions to share and include.