Method/system to identify, extract, and convey key words, phrases, and meanings in service of emergency communications (e.g., voice, text, and video)
AssigneeIntrado Life & Safety, Inc.
InventorsAdan K. Pope, Michel Brkovic
Life of this patent
Estimating the term…
Abstract
A method includes receiving visual content of a physical location, the visual content identified by a content timestamp, the physical location identified by a spatial identifier; receiving audio content of a call; performing voice recognition on the call to extract a first audio symbol; receiving a first timestamp of the call, the first timestamp indicating a time at which the call was initiated or a time extracted from the call by voice recognition; and determining a feature in the visual content, at least in part based on the first audio symbol, the feature defined by a person, object, or situation.
Claims21
1-20. (canceled)
A method, comprising:
- receiving visual content of a physical location, the visual content identified by a content timestamp, the physical location identified by a spatial identifier;
- receiving audio content of a call;
- performing voice recognition on the call to extract a first audio symbol;
- receiving a first timestamp of the call, the first timestamp indicating a time at which the call was initiated or a time extracted from the call by voice recognition;
- identifying the visual content from among a plurality of visual content, at least in part based on the content timestamp and the first timestamp;
- and
- determining a feature in the visual content, at least in part based on the first audio symbol, the feature defined by a person, object, or situation.
An apparatus, comprising:
- a network interface that receives visual content of a physical location and audio content of a call, the visual content identified by a content timestamp, the physical location identified by a spatial identifier;
- and
- a processing unit configured to:
- perform voice recognition on the call to extract a first audio symbol,
- receive a first timestamp of the call, the first timestamp indicating a time at which the call was initiated or a time extracted from the call by voice recognition,
- identify the visual content from among a plurality of visual content, at least in part based on the content timestamp and the first timestamp, and
- determine a feature in the visual content, at least in part based on the first audio symbol, the feature defined by a person, object, or situation.
A method, comprising:
- receiving audio content of a call;
- performing voice recognition on the call to extract a first audio symbol;
- standardizing the first audio symbol to a standardized word based on a thesaurus;
- receiving visual content of a physical location;
- determining a feature in the visual content, at least in part based on the standardized word, the feature defined by a person, object, or situation;
- assigning a relevance score to the first audio symbol;
- and
- contributing the feature and the relevance score to a learning model.
Dependencies are inferred from the claim text.