Speech recognition test method and system based on reverberation simulation

Abstract

The invention provides a voice recognition test method and system based on reverberation simulation. The method comprises the following steps: firstly, setting a first test scene, placing reverberation parameter collection equipment of the first test scene in a first closed boundary and surrounding a preset to-be-tested position; using a reverberation parameter acquisition device to perform reverberation parameter acquisition on a first test audio emitted by a first sound source, then, setting a second test scene, using an audio generator in the second test scene to generate a simulated reverberation test audio, enabling the second sound source to emit the simulated reverberation test audio, and enabling a device to be tested to receive the simulated reverberation test audio; and finally, enabling the device to be tested to output an identification result, and using the processor to judge the identification result. The system comprises the first test scene, the second test scene, the device to be tested and the processor. According to the invention, a method for simulating real reverberation is adopted to replace a traditional method for testing in a real environment, the method is not restricted by sites any more, and the operation is more convenient and faster.

Description

technical field [0001] The present invention relates to the technical field of reverberation simulation, in particular to a method and system for speech recognition testing based on reverberation simulation. Background technique [0002] Language has always been the most important way of communication between people. However, with the rapid development of artificial intelligence, language is no longer just a way of communication between people. Language has also become an important means of communication between people and machines. As a human-machine communication interface, artificial intelligence speech recognition technology has become a key technology for communication between humans and machines. Artificial intelligence speech recognition technology provides various possibilities for our lives, making our lives easier, and it can also To promote innovation in various fields of our lives, there will be opportunities for the application of artificial intelligence speech ...

AI Technical Summary

Problems solved by technology

[0004] Because the artificial intelligence voice equipment will be applied in different reverberation scenarios, the artificial intelligence voice equipment will conduct speech recognition tests in different reverberation scenarios before leaving the factory. Before the voice performance test of the artificial intelligence voice equipment, it is necessary to arrange various Simulating different reverberation conditions in different environments wastes a lot of manpower and material resources, and the layout of various real environments will be restricted by the venue, so some complex scenes are very inconvenient to arrange, and after testing a scene, it is necessary to The test equipment is transferred to another scene for testing, so it is impossible to test multiple scenes at the same time, which is inefficient

Images