/Patent Tracker
SamsungGranted patentAI & softwareInterest 6/10

Samsung patents voice assistant that remembers context across topics

This patent describes a method for electronic devices to handle voice commands that switch between different topics, such as moving from music to navigation. The system saves specific data from the first command to help understand a second command that uses vague references, ensuring the assistant stays on track.

Why it mattersThis improves the natural flow of voice interactions by allowing users to use short, context-dependent phrases without repeating full details.

In the filing’s own words

An electronic device includes: memory storing instructions; and at least one processor. Wherein the instructions which when executed by at least one processor, cause at least one processors to perform control to: receive a first signal corresponding to a first utterance of a user; provide a first response corresponding to the first utterance; obtain cross-domain data based on the first utterance or the first response; receive a second signal corresponding to a second utterance following the first utterance; and provide a second response corresponding to the second utterance based on the cross-domain data when the second utterance includes anaphora. The first utterance and the second utterance respectively correspond to different domains.

What exactly they claim (claim 1)

1. An electronic device comprising: at least one processor including processing circuitry; and memory storing instructions when executed by at the least one processor individually or collectively, cause the electronic device to: receive a first signal corresponding to a first input of a user; provide a first response corresponding to the first input, the first input and the first response corresponding to a first domain; obtain cross-domain data which is used to support a cross-domain scenario in which a consecutive plurality of utterances respectively correspond to different domains, based on the first input or the first response; store the cross-domain data and activation information of the cross-domain data in the memory based on a library domain, the activation information including information on a period during which a voice recognition service activated based on the first input lasts to process an input following the first input; receive a second signal corresponding to a second input of the user following the first input, the second input corresponding to a second domain that is different from the first domain; when the second input is received within the period and includes an anaphora that matches a category of the obtained cross-domain data based on the library domain, read the cross-domain data from the memory, the cross-domain data being obtained from the first input or the first response; and provide a second response corresponding to the second input based on the read cross-domain data when the second input includes the anaphora.

Facts

Company
Samsung (SAMSUNG ELECTRONICS CO., LTD.)
Official title
Electronic device and user utterance processing method
Patent number
US12731584B2
Application number
18587141
Filed
Feb 26, 2024
Granted
Sep 8, 2026
Inventors
5
Claims
15

Classified under

G10L 15/22 Physics & computingG10L 15/19 Physics & computing
Read the full filing (official PDF) ↗ Open in the Patent Tracker

More from Samsung in AI & software

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.