Google patents video matcher that uses sound to find visual similarities
Google describes a method for finding similar video clips by first matching their audio tracks. It then checks if the most visually similar frames line up in a straight diagonal pattern over time, confirming a match.
In the filing’s own words
Audio matching is performed between a first video content item and a second video content item to identify a matching audio segment. Visual similarity values are generated between portions of the first video content item and the second video content item corresponding to the matching audio segment. A top predetermined percentage of highest visual similarity values is selected from the generated visual similarity values. A concentration analysis is performed to determine whether the selected top predetermined percentage of highest visual similarity values are concentrated in a diagonal pattern along a temporal alignment corresponding to the matching audio segment. A match between the first video content item and the second video content item is determined based on the concentration analysis.
What exactly they claim (claim 1)
1. A method, comprising: performing an audio matching between a first video content item and a second video content item to identify a matching audio segment; generating visual similarity values between portions of the first video content item and the second video content item corresponding to the matching audio segment; selecting a top predetermined percentage of highest visual similarity values from the generated visual similarity values; performing a concentration analysis to determine whether the selected top predetermined percentage of highest visual similarity values are concentrated in a diagonal pattern along a temporal alignment corresponding to the matching audio segment; and determining a match between the first video content item and the second video content item based on the concentration analysis.
Facts
- Company
- Google (Google LLC)
- Official title
- Leveraging Audio Matches to Improve Visual Matching Recall Between Video Content Items
- Publication number
- US20260313192A1
- Application number
- 19715367
- Filed
- Jun 22, 2026
- Published
- Oct 8, 2026
- Inventors
- 1
- Claims
- 20
Classified under
H04N 21/4394 Electricity & electronicsG06V 10/761 Physics & computingH04N 21/8106 Electricity & electronicsMore from Google in AI & software
- Google patents video search that answers questions using AI descriptions Sep 15, 2026
- Google patents photo tool that removes people and their gear Sep 15, 2026
- Google patents AI that writes live meeting notes and summaries Sep 15, 2026
- Google patents AI tool that turns messy emergency notes into clear messages Sep 3, 2026
- Google patents AI that clicks through apps to finish tasks for you Sep 24, 2026
Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.