Measuring double talk performance
View Patent ↗A system evaluates a hands free communication system. The system automatically selects a consonant-vowel-consonant (CVC), vowel-consonant-vowel (VCV), or other combination of sounds from an intelligent database. The selection is transmitted with another communication stream that temporally overlaps the selection. The quality of the communication system is evaluated through an automatic speech recognition engine. The evaluation occurs at a location remote from the transmitted selection.
1. A method of evaluating a hands free communication system, comprising:
selecting a consonant-vowel-consonant combination of sounds from an intelligent database automatically;
transmitting the selection when an other communication stream that temporally overlaps the selection is transmitted automatically; and
evaluating the quality of the communication system through an automatic speech recognition engine by measuring the number of correctly recognized words from the selected stream when it is received at a location remote from the transmitted selection.
2. The method of evaluating a hands free communication system of claim 1 where the selection is based on voicing, nasality, sustentation, graveness, or a compactness attribute.
3. The method of evaluating a hands free communication system of claim 1 where the selection is based on voicing, nasality, sustentation, graveness, and a compactness attribute.
4. The method of evaluating a hands free communication system of claim 1 where the consonant-vowel-consonant combination comprises a monosyllable word pair.
5. The method of evaluating a hands free communication system of claim 4 where the monosyllable word pair comprises a selection from a Diagnostic Rhyme Test or a derivative of the Diagnostic Rhyme Test.
6. The method of evaluating a hands free communication system of claim 5 where the monosyllable word pair comprises a plurality of monosyllable word pair that represent voicing, nasality, sustentation, graveness, and a compactness attribute, respectively.
7. The method of evaluating a hands free communication system of claim 1 further comprising measuring a frame of the other communication stream to identify voiced and unvoiced speech segments and processing the measurement to synchronize the transmission of the consonant-vowel-consonant combination with the transmission of the other communication stream.
8. The method of evaluating a hands free communication system of claim 7 further comprising an energy level detector that measures the endpoints of the other communication stream.
9. The method of evaluating a hands free communication system of claim 1 where the other communication stream comprises an artificial voice segment.
10. The method of evaluating a hands free communication system of claim 1 where the other communication stream comprise a synthesis of words from a table of phonemes retained in a memory.
11. The method of evaluating a hands free communication system of claim 1 further comprising a plurality of conversion devices that convert electric signals representing the consonant-vowel-consonant combination and the communication stream into an audible sound.
12. The method of claim 11 further comprising mixing the consonant-vowel-consonant combination audible segment with the other communication stream audible segment with a noise generated from a wideband noise synthesizer.
13. The method of evaluating a hands free communication system of claim 1 further comprising transmitting the quality evaluation to a remote data analysis controller through a publicly accessible distributed network.
14. A system that evaluates a hands free communication system, comprising:
a audio database management system that retains voiced and/or unvoiced artificial voice or speech segments to be transmitted through a loudspeaker;
a voice activity detector configured to measure a frame of the voiced and/or unvoiced artificial voice or speech segment;
a synchronizer that compensates for differences in rates of communication between the audio database management system and a separate server that accesses diagnostic word pairs;
a narrowband noise synthesizer configured to generate and transmit periodic and/or non-periodic noise signals temporally with the output of the loudspeaker;
a mixer that receives and mixes the output of the loudspeaker, an audible representation of the diagnostic word pairs, and an output of the narrowband noise synthesizer;
an automatic speech recognition engine in communication with the mixer programmed to select one of the diagnostic word pairs; and
a data analysis controller remote from the automatic speech recognition engine configured to measure the level of intelligibility of a hands free system based on an evaluation of the automatic speech recognition engine's word selections.
15. The system that evaluates a hands free communication system of claim 14 where the data analysis controller is resident to an event-driven-server.
16. The system that evaluates the hands free communication system of claim 14 where the data analysis controller is resident to a process-server.
17. The system that evaluates the hands free communication system of claim 16 where the process-server renders an impression characteristic that is transmitted to a client device.
18. The system that evaluates the hands free communication system of claim 17 further comprising a graphical user interface configured to present the diagnostic word pairs to a user and receive the selection of the two words that the user perceived to have heard.
19. The system that evaluates the hands free communication system of claim 14 further comprising a first convolver interfaced to the audio database management system and a second convolver interfaced to the server.
20. A system that evaluates a hands free communication system, comprising:
a audio database management system that retains voiced and/or unvoiced artificial voice or speech segment to be transmitted through a loudspeaker;
a voice activity detector configured to measure a frame of the voiced and/or unvoiced artificial voice or speech segment;
a synchronizer that compensates for differences in rates of communication between the audio database management system and a separate server that accesses diagnostic word pairs;
a narrowband noise synthesizer configured to generate and transmit periodic and/or non-periodic noise signals temporally with the output of the loudspeaker;
a mixer that receives and mixes the output of the loudspeaker, an audible representation of the diagnostic word pairs, and an output of the narrowband noise synthesizer;
an automatic speech recognition engine in communication with the mixer programmed to select one of the diagnostic word pairs; and
means remote from the automatic speech recognition engine configured to measure the level of intelligibility of a hands free system based on an evaluation of an automatic speech recognition engine's word selections and a user's word selection received through a user interface.