Voice assistant error detection system
A vehicle system for classifying spoken utterance within a vehicle cabin as one of system-directed and non-system directed, the system may include at least one microphone configured to detect at least one acoustic utterance from at least one occupant of a vehicle, at least one sensor to detect user behavior data indicative of user behavior, and a processor programmed to: receive the acoustic utterance, classify the acoustic utterance as one of a system-directed utterance and a non-system directed utterance, determine whether the acoustic utterance was properly classified based on user behavior observed via data received from the sensor after the classification, and apply a mitigating adjustment to classifications of subsequent acoustic utterances based on an improper classification.
1. A vehicle system for classifying spoken utterance within a vehicle cabin as one of system-directed and non-system directed, the system comprising:
at least one microphone configured to detect at least one acoustic utterance from at least one occupant of a vehicle;
at least one sensor to detect user behavior data indicative of user behavior; and
a processor programmed to:
receive the acoustic utterance,
classify the acoustic utterance as one of a system-directed utterance and a non-system directed utterance,
determine whether the acoustic utterance was properly classified based on user behavior observed via data received from the sensor after the classification, and
apply a mitigating adjustment to classifications of subsequent acoustic utterances based on an improper classification, wherein the mitigating adjustment includes adjusting a confidence threshold for determining whether the acoustic utterance is one of a system-directed utterance and a non-system directed utterance.
2. The system of claim 1 , wherein the processor is further programmed to transmit and receive the mitigating adjustment from an off-board server.
3. The system of claim 1 , wherein the user behavior includes repeating of the acoustic utterance.
4. The system of claim 1 , wherein the user behavior includes a corrective command.
5. The system of claim 1 , wherein the user behavior includes a subsequent acoustic utterance with differing pronunciation or tone than that of the prior acoustic utterance.
6. The system of claim 1 , wherein the user behavior includes an action indicative of canceling of the acoustic utterance.
7. The system of claim 1 , wherein the user behavior includes a subsequent acoustic utterance including a profanity.
8. A method for classifying spoken utterances within a vehicle cabin as one of system-directed and non-system directed, the method comprising:
receiving an acoustic utterance;
classifying the acoustic utterance as one of a system-directed utterance and a non-system directed utterance;
determining whether the acoustic utterance was properly classified based on user behavior observed via data received from a sensor after the classification; and
applying a mitigating adjustment to classifications of subsequent acoustic utterances based on an improper classification, wherein the mitigating adjustment includes identifying an error specific to a semantic domain and ignoring classification of that domain.
9. The method of claim 8 , further comprising transmitting and receiving the mitigating adjustment from an off-board server.
10. The method of claim 8 , wherein the user behavior includes repeating of the acoustic utterance.
11. The method of claim 8 , wherein the user behavior includes a corrective command.
12. The method of claim 8 , wherein the user behavior includes a subsequent acoustic utterance with differing pronunciation or tone than that of the prior acoustic utterance.
13. The method of claim 8 , wherein the user behavior includes an action indicative of canceling of the acoustic utterance.
14. A non-transitory computer-readable medium storing computer-executable instructions, that when executed by at least one processor, cause the at least one processor to:
receive an acoustic utterance,
classify the acoustic utterance as one of a system-directed utterance and a non-system directed utterance,
determine whether the acoustic utterance was properly classified based on user behavior observed via data received from a sensor after the classification, and
apply a mitigating adjustment to classifications of subsequent acoustic utterances based on an improper classification, wherein the mitigating adjustment includes identifying an error specific to a dialog and situation and ignoring classification during the situation.