US 12,614,543 B1Granted

Systems and methods for on-device spoken language understanding

AssigneeAmazon Technologies, Inc.

InventorsAnastasios Alexandridis, Kanthashree Mysore Sathyendra, Grant Strimel, Pavel Kveton, Jon A. Webb, Ariya Rastrow

Life of this patent

Estimating the term…

patentgrep

Loading patent…

Abstract

Techniques for performing spoken language understanding (SLU) processing on a device are described. Example embodiments involve a device determining whether a spoken input corresponds to a supported spoken input class, a supported spoken input with dynamic content class or an unsupported spoken input class. For a spoken input corresponding to the supported spoken input with dynamic content class, the device may determine an entity corresponding to the spoken input from a set of entities, which may be determined based on device context data and/or user profile data. For a spoken input corresponding to the supported spoken input class, the device may determine an intent and entity using stored data. For a spoken input corresponding to the unsupported spoken input class, the device may send the audio data to a system for processing.

Claims20

Dependent claims

A computer-implemented method comprising:

receiving, at a device, first audio data corresponding to a first spoken input;
determining, at the device and using a spoken language understanding (SLU) component, that the first audio data corresponds to a first category representing a first supported spoken input with dynamic content, wherein the SLU component is configured to classify the first audio data to:
the first category,
a second category representing a first supported spoken input, and
a third category representing an unsupported spoken input;
based on the first audio data corresponding to the first category, determining a first plurality of entities potentially corresponding to the first spoken input;
processing, at the device and using the SLU component, the first audio data with respect to the first plurality of entities to:
determine a portion of the first audio data corresponding to a mention of an entity, and
determine that the portion of the first audio data represents a first entity of the first plurality of entities;
and
determining, at the device, a first intent corresponding to the first category.

8 more elements

A system comprising:

at least one processor;
and
at least one memory including instructions that, when executed by the at least one processor, cause the system to:
receive first audio data corresponding to a first spoken input;
determine, using a spoken language understanding (SLU) component, that the first audio data corresponds to a first category representing a first supported spoken input with dynamic content, wherein the SLU component is configured to classify the first audio data to:
the first category,
a second category representing a first supported spoken input, and
a third category representing an unsupported spoken input;
based on the first audio data corresponding to the first category, determine a first plurality of entities potentially corresponding to the first spoken input;
process, using the SLU component, the first audio data with respect to the first plurality of entities to:
determine a portion of the first audio data corresponding to a mention of an entity, and
determine that the portion of the first audio data represents a first entity of the first plurality of entities;
and
determine a first intent corresponding to the first category.

11 more elements

Dependencies are inferred from the claim text.