Voice control in a multi-talker and multimedia environment

This invention describes a system for understanding voice commands in a room where multiple people might be talking and various media (like TV or music) are playing. It works by placing microphones in different zones, then cleaning up the sound from each zone by removing echoes from speakers and filtering out sounds from other zones. Finally, it performs speech recognition on these cleaned signals to understand what is being said in the active zone.

Why it matters: Filed before the widespread integration of large language models and advanced deep learning into speech recognition. These advancements have significantly improved robustness in adverse environments and semantic understanding, making the core challenges more tractable.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedHeavy software system, complex signal processing

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10