US Patent Application 18218818. USING CORRECTIONS, OF AUTOMATED ASSISTANT FUNCTIONS, FOR TRAINING OF ON-DEVICE MACHINE LEARNING MODELS simplified abstract
USING CORRECTIONS, OF AUTOMATED ASSISTANT FUNCTIONS, FOR TRAINING OF ON-DEVICE MACHINE LEARNING MODELS
Organization Name
Inventor(s)
Françoise Beaufays of Mountain View CA (US)
Rajiv Mathews of Sunnyvale CA (US)
Dragan Zivkovic of Sunnyvale CA (US)
Kurt Partridge of San Francisco CA (US)
Andrew Hard of Menlo Park CA (US)
USING CORRECTIONS, OF AUTOMATED ASSISTANT FUNCTIONS, FOR TRAINING OF ON-DEVICE MACHINE LEARNING MODELS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18218818 titled 'USING CORRECTIONS, OF AUTOMATED ASSISTANT FUNCTIONS, FOR TRAINING OF ON-DEVICE MACHINE LEARNING MODELS
Simplified Explanation
The patent application describes a system where a client device can use sensor data and machine learning to activate automated assistant functions. If the device makes an incorrect decision, it can generate a gradient to update the on-device speech recognition model. The gradient can also be sent to a remote system to update a global speech recognition model.
- Client devices can use sensor data and machine learning to activate automated assistant functions.
- If the device makes an incorrect decision, it can generate a gradient based on the predicted output compared to the ground truth output.
- The generated gradient can be used to update the on-device speech recognition model.
- The gradient can also be sent to a remote system to update a global speech recognition model.
Original Abstract Submitted
Processor(s) of a client device can: receive sensor data that captures environmental attributes of an environment of the client device; process the sensor data using a machine learning model to generate a predicted output that dictates whether one or more currently dormant automated assistant functions are activated; making a decision as to whether to trigger the one or more currently dormant automated assistant functions; subsequent to making the decision, determining that the decision was incorrect; and in response to determining that the determination was incorrect, generating a gradient based on comparing the predicted output to ground truth output. In some implementations, the generated gradient is used, by processor(s) of the client device, to update weights of the on-device speech recognition model. In some implementations, the generated gradient is additionally or alternatively transmitted to a remote system for use in remote updating of global weights of a global speech recognition model.