Back to results
US 12,333,611 B1Granted
Digital imagery, audio, and meta-data
AssigneeUnited Services Automobile Association (USAA)
InventorsGregory David Hansen, Andre R. Buentello
Life of this patent
Estimating the term…
Abstract
Methods, systems, and apparatus, including computer programs encoded on computer storage media, to process audio and visual data. One of the methods includes receiving, from a mobile device, image data and audio data, wherein the image data and audio data were obtained within the same time frame. The method includes processing the image data to identify a first set of facts. The method includes processing the audio data to identify a second set of facts. The method includes processing meta-data associated with the image data to identify a third set of facts. The method also includes performing at least one process using the first set of facts, the second set of facts, and the third set of facts.
Claims18
A system, comprising:
- a memory configured to store instructions;
- and
- one or more processors communicatively coupled to the memory, wherein the one or more processors are configured to execute the instructions and perform operations comprising:
- receiving, from a mobile device, a data package comprising image data and audio data, wherein the image data and the audio data are acquired during the same time period by the mobile device, and wherein the audio data comprises speech describing a set of facts regarding an accident, and one or more weather conditions;
- extracting metadata from the image data and the audio data, wherein the metadata comprises one or more time stamps and global positioning system (GPS) data;
- identifying and detecting one or more objects and one or more object conditions within images, wherein the one or more object conditions comprises the one or more weather conditions and one or more road conditions;
- deriving weather data based on the extracted metadata and the image data;
- determining, via one or more machine learning algorithms, whether the image data comprises the one or more objects and the one or more object conditions representative of one or more components of the set of facts described in the speech, wherein the one or more objects comprise rain, snow, ice, or any combination thereof;
- and
- authenticating an insurance claim based on the derived weather data corroborating the set of facts using speech recognition, the image data, and the metadata to confirm that the one or more objects and the one or more object conditions corroborate the weather data.
7 more elements
A method, comprising:
- receiving, via a processing system, a data package comprising image data and audio data, wherein the audio data comprises speech describing a set of components, wherein the image data and the audio data are acquired by a mobile device during the same time period, and wherein the audio data comprises speech describing a set of facts regarding an accident and one or more weather conditions;
- extracting, via the processing system, metadata from the image data and the audio data, wherein the metadata comprises one or more time stamps and global positioning systems (GPS) data;
- identifying and detecting, via the processing system, one or more objects and one or more object conditions within images, wherein the one or more object conditions comprises the one or more weather conditions and one or more road conditions;
- deriving, via the processing system, weather data based on the extracted metadata and the image data;
- determining, via one or more machine learning algorithms, whether the image data comprises the one or more objects and the one or more object conditions representative of one or more components of the set of facts described in the speech, wherein the one or more objects comprise rain, snow, ice, or any combination thereof;
- and
- authenticating, via the processing system, an insurance claim in response to the derived weather data corroborating the set of facts using speech recognition, the image data, and the metadata to confirm that the one or more objects and the one or more object conditions corroborate the weather data.
4 more elements
A non-transitory computer storage medium comprising computer-executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
- receiving, from a mobile device, a data package comprising image data and audio data, wherein the audio data comprises speech describing a set of facts regarding an accident and one or more weather conditions, wherein the image data and the audio data are acquired by the mobile device during the same time period;
- extracting metadata from the image data and the audio data, wherein the metadata comprises one or more time stamps and global positioning system (GPS) data;
- identifying and detecting one or more objects and one or more object conditions within images, wherein the one or more object conditions comprises the one or more weather conditions and one or more road conditions;
- deriving weather data based on the extracted metadata and the image data;
- determining, via one or more machine learning algorithms, whether the image data comprises the one or more objects and the one or more objects conditions representative of one or more components of the set of facts described in the speech, wherein the one or more objects comprise rain, snow, ice, or any combination thereof;
- and
- authenticating an insurance claim based on the derived weather data corroborating the set of facts using speech recognition, the image data, and the metadata to confirm that the one or more objects and the one or more object conditions corroborate the weather data.
4 more elements
Dependencies are inferred from the claim text.