2021 6th International Conference on Smart Grid and Electrical Automation (ICSGEA) | 2021
Voiceprint Recognition Based on Big Data and Gaussian Mixture Model
Abstract
In order to overcome the disadvantages of voiceprint recognition technology in application, for example, it is difficult to accurately identify the voice of different groups of people, it is difficult to accurately label the voice of each individual, and so on, this paper proposes a novel voiceprint recognition technology based on big data technology and Gaussian mixture model. The technology adopts big data technology, introduces big data thinking, gives full play to the processing and storage capacity of big data technology, and can quickly process and store different sound data. Based on this, Gaussian mixture model is used to analyze the voiceprint differences of different individuals, and EM model is used to label different samples accurately. At the same time, the technology also uses Matlab algorithm to realize the optimal strategy of identifying individual voice differences. The research results show that, different from the traditional voiceprint recognition technology, the new voiceprint recognition technology proposed in this paper can achieve 95% accuracy whether recording or loading.