Non-speech input to speech processing system
A system and method for incorporating motion into a speech processing system. A wearable device that is capable of both capturing spoken utterances and capturing motion data may be used to interact with a speech processing system. In certain circumstances, such as when voice communication are unreliable (due to noise) or when controlling the system by motion is desired, motion of a device may be used to provide input to a speech processing system. For example, sensor data or gesture data resulting from movement of a device may be processed and input into a natural language system as representative of a spoken command portion or other input. The motion information may be interpreted to provide prompts to the system (e.g., “yes,” “no,” etc.), to perform certain commands (skip, forward, back, cancel) or to otherwise control the system.
1. A computer-implemented method of using motion data to interact with a speech processing system, the method comprising:
receiving input audio data from a wearable device;
associating the input audio data with an indicator;
performing automatic speech processing (ASR) on the input audio data to determine first text data;
performing natural language understanding (NLU) processing on the first text data to determine a command;
determining that execution of the command requires further input from a user;
determining prompt text data corresponding to a solicitation of the further input;
performing text-to-speech (TTS) processing on the prompt text data to determine prompt audio data;
sending the prompt audio data to the wearable device;
receiving further data from the wearable device, the further data corresponding to rotation of the wearable device about at least one axis;
associating the further data with the indicator;
processing the further data to determine that the further data corresponds to the command;
determining that the further data satisfies a condition;
performing further NLU processing using the further data and the indicator to determine a response to the prompt audio data; and
executing the command based at least in part on the response.
2. The computer-implemented method of claim 1 , wherein:
determining the prompt text data comprises determining text data corresponding to a request for the user to answer “yes” or “no”;
receiving the further data comprises receiving an indication that a user nodded; and
the method further comprises determining the user answered “yes” based at least in part on the further data.
3. The computer-implemented method of claim 1 , wherein the further data comprises sensor data from at least one motion sensor of the wearable device, and wherein the method further comprises:
comparing the sensor data to stored data to identify a physical gesture performed with the wearable device that resulted in the sensor data; and
determining response data corresponding to the physical gesture.
4. A system comprising:
at least one processor; and
memory including instructions operable to be executed by the at least one processor to perform a set of actions to configure the at least one processor to:
receive, from a first device, input audio data corresponding to an utterance;
perform automatic speech recognition (ASR) on the input audio data to determine text data;
determine a command potentially corresponding to the text data;
determine processing of the command requires further input;
receive, from the first device, motion data corresponding to the input audio data the motion data representing a rotation of the first device about at least one axis;
process the motion data to determine that the motion data corresponds to the command;
determine that the motion data satisfies a condition; and
use the motion data to process the command.
5. The system of claim 4 , wherein the at least one processor is further configured to determine natural language understanding output based at least in part on the text data and the motion data.
6. The system of claim 4 , wherein the at least one processor is further configured to:
perform natural language understanding on the text data to determine the command;
determine prompt text data corresponding to a solicitation of the further input;
perform text-to-speech (TTS) processing on the prompt text data to determine prompt audio data; and
send the prompt audio data to the first device,
wherein the motion data is received after the prompt audio data is sent.
7. The system of claim 4 , wherein the first device is a wearable device comprising at least one motion sensor to detect a first user movement, the motion data generated based on the first user movement.
8. The system of claim 7 , wherein the motion data comprises an indication that a physical gesture was performed using the first device.
9. The system of claim 8 , wherein the at least one processor is further configured to:
identify stored text data corresponding to the physical gesture; and
process the stored text data using a natural language understanding component to determine an intent.
10. The system of claim 9 , wherein the at least one processor is further configured to identify a command corresponding to the sensor data.
11. The system of claim 7 , wherein the motion data comprises sensor data created by at least one motion sensor of the first device.
12. The system of claim 4 , wherein the motion data comprises an indication of a command associated with a first user movement.
13. The system of claim 4 , wherein the at least one processor is further configured to:
determine the input audio data and the motion data correspond to a first time period.
14. A computer-implemented method comprising:
receiving, from a first device, input audio data corresponding to an utterance;
performing automatic speech recognition (ASR) on the input audio data to determine text data;
determining a command potentially corresponding to the text data;
determining that processing of the command requires further input;
receiving, from the first device, motion data corresponding to the input audio data, the motion data representing a rotation of the first device about at least one axis;
processing the motion data to determine that the motion data corresponds to the command;
determining that the motion data satisfies a condition; and
using the motion data to process the command.
15. The computer-implemented method of claim 14 , further comprising determining natural language understanding output based at least in part on the text data and the motion data.
16. The computer-implemented method of claim 14 , further comprising:
performing natural language understanding on the text data to determine the command;
determining prompt text data corresponding to a solicitation of the further input;
performing text-to-speech (TTS) processing on the prompt text data to determine prompt audio data; and
sending the prompt audio data to the first device,
wherein the motion data is received after the prompt audio data is sent.
17. The computer-implemented method of claim 14 , wherein the first device is a wearable device comprising at least one motion sensor to detect a first user movement, the motion data generated based on the first user movement.
18. The computer-implemented method of claim 17 , wherein the motion data comprises an indication that a physical gesture was performed using the first device.
19. The computer-implemented method of claim 18 , further comprising:
identifying stored text data corresponding to the physical gesture; and
processing the stored text data using a natural language understanding component to determine an intent.
20. The computer-implemented method of claim 17 , wherein the motion data comprises sensor data created by at least one motion sensor of the first device.
21. The computer-implemented method of claim 14 , wherein the motion data comprises an indication of a command associated with a first user movement.
22. The computer-implemented method of claim 14 , further comprising determining the input audio data and the motion data correspond to a first time period.