Audio analysis for text generation
OwnerRon Zass, Ben Avi Ingelinventors
Life of this patent
Estimating the term…
Abstract
Systems, methods and non-transitory computer readable media for analyzing audio data for text generation are provided. Audio data captured using at least one audio sensor may be received. The audio data may be analyzed to identify a plurality of objects. For each object of the plurality of objects, data associated with the object may be analyzed to select an adjective, and a description of the object that includes the adjective may be generated. Further, a textual content that includes the generated descriptions of the plurality of objects may be generated. The generated textual content may be provided.
Claims21
A non-transitory computer readable medium storing a software program comprising data and computer implementable instructions that when executed by at least one processor cause the at least one processor to perform operations for analyzing audio data to generate a textual content reporting objects, the operations comprising:
- receiving audio data captured using at least one audio sensor;
- calculating a convolution of at least part of the audio data and thereby obtain result values of the calculated convolution;
- identifying a plurality of objects based on the result values;
- obtaining an indication of a writer persona;
- for each object of the plurality of objects, analyzing respective pitch values of respective at least part of the audio data to determine data associated with the object;
- for each object of the plurality of objects, analyzing the data associated with the object to determine a respective mathematical object in a mathematical space, using the indication of the writer persona and the respective mathematical object to select a respective adjective, and generating a description of the object that includes the respective adjective in a style associated with the writer persona, the style includes an affinity to usage of specific words, the style is associated with a degree of sarcasm;
- using a generative model to generate a textual content that includes the generated descriptions of the plurality of objects;
- and
- providing the generated textual content.
A system for analyzing audio data to generate a textual content reporting objects, the system comprising:
- at least one processing unit configured to perform the operations of:
- receiving audio data captured using at least one audio sensor;
- calculating a convolution of at least part of the audio data and thereby obtain result values of the calculated convolution;
- identifying a plurality of objects based on the result values;
- obtaining an indication of a writer persona;
- for each object of the plurality of objects, analyzing respective pitch values of respective at least part of the audio data to determine data associated with the object;
- for each object of the plurality of objects, analyzing the data associated with the object to determine a respective mathematical object in a mathematical space, using the indication of the writer persona and the respective mathematical object-to select a respective adjective, and generating a description of the object that includes the respective adjective in a style associated with the writer persona, the style includes an affinity to usage of specific words, the style is associated with a degree of sarcasm;
- using a generative model to generate a textual content that includes the generated descriptions of the plurality of objects;
- and
- providing the generated textual content.
A method for analyzing audio data to generate a textual content reporting objects, the method comprising:
- receiving audio data captured using at least one audio sensor;
- calculating a convolution of at least part of the audio data and thereby obtain result values of the calculated convolution;
- identifying a plurality of objects based on the result values;
- obtaining an indication of a writer persona;
- for each object of the plurality of objects, analyzing respective pitch values of respective at least part of the audio data to determine data associated with the object;
- for each object of the plurality of objects, analyzing the data associated with the object to determine a respective mathematical object in a mathematical space, using the indication of the writer persona and the respective mathematical object to select a respective adjective, and generating a description of the object that includes the respective adjective in a style associated with the writer persona, the style includes an affinity to usage of specific words, the style is associated with a degree of sarcasm;
- using a generative model to generate a textual content that includes the generated descriptions of the plurality of objects;
- and
- providing the generated textual content.
Dependencies are inferred from the claim text.