Equipment control method and device, storage medium and program product

Abstract

The embodiment of the invention discloses an equipment control method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining equipment control data; identifying the data type of the equipment control data; based on the data type of the equipment control data, respectively inputting data of each data type in the equipment control data into a corresponding identification model in a multi-model system, and outputting an intermediate identification result through the corresponding identification model; wherein the multi-model system comprises a plurality of recognition models, and any two recognition models in the plurality of recognition models are obtained by training based on different modes; and determining a final identification result corresponding to the equipment control data based on the intermediate identification result corresponding to the data of each data type in the equipment control data so as to control the electronic equipment based on the final identification result. According to the embodiment of the invention, the control accuracy of the electronic equipment can be improved.

Description

technical field [0001] The present disclosure relates to artificial intelligence technology, and in particular, to a device control method and device, a storage medium and a program product. Background technique [0002] Speech recognition technology has made significant progress in recent years, and has now entered various fields such as industry, home appliances, and smart homes. Voice wake-up is a form of voice recognition technology. It does not directly contact the hardware device, and the device can be woken up and run by the voice of the keyword. In daily life, many electronic devices have voice wake-up function. [0003] For the voice wake-up function, there are two opposite key indicators: wake-up rate and false wake-up rate. In the actual speech environment, due to various complex signal sources, noise and other interferences, and similar pronunciations, the electronic device may be awakened by mistake, or the electronic device may not be able to be timely due to...

AI Technical Summary

Problems solved by technology

In the actual voice environment, due to the presence of various complex signal sources, noise and other interference, and similar pronunciation, it may cause the electronic device to be awakened by mistake, or the electronic device cannot be activated in time due to noise interference after the user speaks the wake-up word. response

[0004] In the process of realizing the present disclosure, the inventors found through research that currently, a single voice model is used to wake up the device by voice. Some voice models have a higher wake-up rate but a high false wake-up rate. The wake-up rate is low and the wake-up rate is also low, and the above two key indicators cannot be taken into account at the same time, resulting in poor wake-up effect

Images