Voice-based age recognition method and device

Abstract

The invention discloses a voice-based age recognition method. The method comprises the steps: acquiring an original audio, carrying out early-stage processing, and determining a training data set; training a selected neural network model by using the training data set, and determining an algorithm model for age recognition; and after acquisition of real-time audio data, performing identification processing on the audio data through the algorithm model, and determining the age of a speaker. The invention further discloses a voice-based age recognition device. According to the scheme disclosed by the invention, the age of the speaker can be recognized based on the user voice, so that the recognition accuracy is high, the user interaction experience is improved, and the requirements of different scenes are met.

Description

technical field [0001] The invention relates to the field of intelligent voice technology, in particular to a voice-based age recognition method and device. Background technique [0002] In recent years, with the development of intelligent voice technology, products based on intelligent voice emerge in an endless stream. In order to improve user experience, these voice products are developing different application modes based on age needs. Based on this, how to accurately identify the age level of the object in the user dialogue is very important. At present, the more mature age recognition technologies on the market are based on face recognition, and the voice-based age recognition technology is still in the research stage, and some implementation solutions have been proposed in the industry, but these existing solutions generally have the following problems to be overcome: [0003] First, the required test audio is too long, and the tester needs to speak continuously for...

AI Technical Summary

Problems solved by technology

[0003] First, the required test audio is too long, and the tester needs to speak continuously for 5 seconds, so a sentence needs to be more than 20 words, which is very difficult for users, resulting in poor user experience and is not conducive to the landing of the product;

[0004] Second, these technologies do not perform well in noisy scenes such as vehicles and background TVs

Images