Preserving audio data collection privacy in mobile devices

Abstract

Techniques are disclosed for using the hardware and/or software of the mobile device to obscure speech in the audio data before a context determination is made by a context awareness application using the audio data. In particular, a subset of a continuous audio stream is captured such that speech (words, phrases and sentences) cannot be reliably reconstructed from the gathered audio. The subset is analyzed for audio characteristics, and a determination can be made regarding the ambient environment.

1 . (canceled) 
     
     
         2 . A method for performing an audio analysis, the method comprising:
 receiving, by a computerized device, a continuous audio stream;   capturing, by the computerized device, from the continuous audio stream, a plurality of audio frames from a plurality of audio blocks of the continuous audio stream, wherein:
 each audio block of the plurality of audio blocks includes multiple audio frames; and 
 a single audio frame is captured from each audio block of the plurality of audio blocks; 
   analyzing, by the computerized device, the plurality of audio frames; and   determining, based on analyzing the plurality of audio frames, a characteristic of an ambient environment of the continuous audio stream.   
     
     
         3 . The method for performing the audio analysis of  claim 2 , wherein the continuous audio stream comprises human speech. 
     
     
         4 . The method for performing the audio analysis of  claim 3 , the method further comprising:
 determining, by the computerized device, based on analyzing the plurality of audio frames, an identity of a speaker of the human speech.   
     
     
         5 . The method for performing the audio analysis of  claim 2 , the method further comprising:
 shuffling, by the computerized device, the plurality of audio frames into a shuffled order, wherein analyzing the plurality of audio frames comprises analyzing the plurality of audio frames in the shuffled order.   
     
     
         6 . The method for performing the audio analysis of  claim 2 , the method further comprising:
 for each audio frame, randomizing, by the computerized device, selection of the audio frame from the multiple audio frames present within the corresponding audio block of the plurality of audio blocks.   
     
     
         7 . The method for performing the audio analysis of  claim 6 , wherein randomizing selection of the audio frame is based, at least in part, on information selected from a source of the group comprising:
 a global navigation satellite system (GNSS) device,   signal noise from circuitry within a mobile device,   signal noise from a microphone, and   signal noise from an antenna.   
     
     
         8 . The method for performing the audio analysis of  claim 2 , further comprising:
 uploading, by the computerized device, the plurality of audio frames to a remote server system, wherein determining, based on analyzing the plurality of audio frames, the characteristic of the ambient environment of the continuous audio stream is performed by the remote server system.   
     
     
         9 . The method for performing the audio analysis of  claim 2 , wherein receiving the continuous audio stream occurs via a microphone of the computerized device and the computerized device is a cellular phone. 
     
     
         10 . A system for performing an audio analysis, the system comprising:
 one or more processors; and   a memory communicatively coupled with and readable by the one or more processors and having stored therein processor-readable instructions which, when executed by the one or more processors, cause the one or more processors to:
 capture, from a continuous audio stream, a plurality of audio frames from a plurality of audio blocks of the continuous audio stream, wherein:
 each audio block of the plurality of audio blocks includes multiple audio frames; and 
 a single audio frame is captured from each audio block of the plurality of audio blocks; 
 
 analyze the plurality of audio frames; and 
 determine, based on analyzing the plurality of audio frames, a characteristic of an ambient environment of the continuous audio stream. 
   
     
     
         11 . The system for performing the audio analysis of  claim 10 , wherein the continuous audio stream captured by the processor comprises human speech. 
     
     
         12 . The system for performing the audio analysis of  claim 11 , wherein the processor-readable instructions, when executed, further cause the one or more processors to:
 determine, based on analyzing the plurality of audio frames, an identity of a speaker of the human speech.   
     
     
         13 . The system for performing the audio analysis of  claim 10 , wherein the processor-readable instructions, when executed, further cause the one or more processors to:
 shuffle the plurality of audio frames into a shuffled order, wherein analyzing the plurality of audio frames comprises analyzing the plurality of audio frames in the shuffled order.   
     
     
         14 . The system for performing the audio analysis of  claim 10 , wherein the processor-readable instructions, when executed, further cause the one or more processors to:
 for each audio frame, randomize selection of the audio frame from the multiple audio frames present within the corresponding audio block of the plurality of audio blocks.   
     
     
         15 . The system for performing the audio analysis of  claim 14 , wherein the processor-readable instructions that, when executed, cause the one or more processors to randomize selection of the audio frame bases the randomization, at least in part, on information selected from a source of the group comprising:
 a global navigation satellite system (GNSS) device,   signal noise from circuitry within a mobile device,   signal noise from a microphone, and   signal noise from an antenna.   
     
     
         16 . The system for performing the audio analysis of  claim 10 , wherein the system is implemented as part of a cellular phone comprising a microphone. 
     
     
         17 . A non-transitory processor-readable medium for performing an audio analysis, comprising processor-readable instructions configured to cause one or more processors to:
 capture, from a continuous audio stream, a plurality of audio frames from a plurality of audio blocks of the continuous audio stream, wherein:
 each audio block of the plurality of audio blocks includes multiple audio frames; and 
 a single audio frame is captured from each audio block of the plurality of audio blocks; 
   analyze the plurality of audio frames; and   determine, based on analyzing the plurality of audio frames, a characteristic of an ambient environment of the continuous audio stream.   
     
     
         18 . The non-transitory processor-readable medium for performing the audio analysis of  claim 17 , wherein the continuous audio stream captured by the processor comprises human speech. 
     
     
         19 . The non-transitory processor-readable medium for performing the audio analysis of  claim 18 , wherein the processor-readable instructions are further configured to cause the one or more processors to:
 determine, based on analyzing the plurality of audio frames, an identity of a speaker of the human speech.   
     
     
         20 . The non-transitory processor-readable medium for performing the audio analysis of  claim 17 , wherein the processor-readable instructions are further configured to cause the one or more processors to:
 shuffle the plurality of audio frames into a shuffled order, wherein analyzing the plurality of audio frames comprises analyzing the plurality of audio frames in the shuffled order.   
     
     
         21 . The non-transitory processor-readable medium for performing the audio analysis of  claim 17 , wherein the processor-readable instructions are further configured to cause the one or more processors to:
 for each audio frame, randomize selection of the audio frame from the multiple audio frames present within the corresponding audio block of the plurality of audio blocks.   
     
     
         22 . The non-transitory processor-readable medium for performing the audio analysis of  claim 21 , wherein the processor-readable instructions configured to cause the one or more processors to randomize selection of the audio frame bases the randomization, at least in part, on information selected from a source of the group comprising:
 a global navigation satellite non-transitory processor-readable medium (GNSS) device,   signal noise from circuitry within a mobile device,   signal noise from a microphone, and   signal noise from an antenna.   
     
     
         23 . The non-transitory processor-readable medium for performing the audio analysis of  claim 17 , wherein the non-transitory processor-readable medium is implemented as part of a cellular phone comprising a microphone. 
     
     
         24 . An apparatus for performing an audio analysis, the apparatus comprising:
 means for receiving a continuous audio stream;   means for capturing from the continuous audio stream, a plurality of audio frames from a plurality of audio blocks of the continuous audio stream, wherein:
 each audio block of the plurality of audio blocks includes multiple audio frames; and 
 a single audio frame is captured from each audio block of the plurality of audio blocks; 
   means for analyzing the plurality of audio frames; and   means for determining, based on analyzing the plurality of audio frames, a characteristic of an ambient environment of the continuous audio stream.   
     
     
         25 . The apparatus for performing the audio analysis of  claim 24 , wherein the continuous audio stream comprises human speech. 
     
     
         26 . The apparatus for performing the audio analysis of  claim 25 , the apparatus further comprising:
 means for determining, based on analyzing the plurality of audio frames, an identity of a speaker of the human speech.   
     
     
         27 . The apparatus for performing the audio analysis of  claim 24 , the apparatus further comprising:
 means for shuffling the plurality of audio frames into a shuffled order, wherein analyzing the plurality of audio frames comprises analyzing the plurality of audio frames in the shuffled order.   
     
     
         28 . The apparatus for performing the audio analysis of  claim 24 , the apparatus further comprising:
 means for randomizing, for each frame, selection of the audio frame from the multiple audio frames present within the corresponding audio block of the plurality of audio blocks.   
     
     
         29 . The apparatus for performing the audio analysis of  claim 28 , wherein the means for randomizing selection of the audio frame bases randomization, at least in part, on information selected from a source of the group consisting of:
 a global navigation satellite system (GNSS) device,   signal noise from circuitry within a mobile device,   signal noise from a microphone, and   signal noise from an antenna.   
     
     
         30 . The apparatus for performing the audio analysis of  claim 24 , further comprising:
 means for uploading the plurality of audio frames to a remote server system, wherein the means for determining, based on analyzing the plurality of audio frames, the characteristic of the ambient environment of the continuous audio stream is present at the remote server system.   
     
     
         31 . The apparatus for performing the audio analysis of  claim 24 , wherein the apparatus is integrated as part of a cellular phone.