Research on Interactive English Speech Recognition Algorithm in Multimedia Cooperative Teaching

摘要

Multi-media overcomes the defects of traditional teaching means so foreign language teaching rapidly develops with such technology. It becomes a bottleneck to restrict intelligence learning software development. To solve the problem, this paper discuss basic knowledge in speech recognition and studies targeted corpus according to English pronunciation habit of Chinese people. Integrated with oral English learners' requirements with Chinese as native language, this paper applies DTW model-based speech recognition technology for Viterbi decoding speech, then it recognizes and scores through posterior probability. After experiment verification, English pronunciation recognition model in this paper is verified to be reasonable and credible and it can offer learners' timely, accurate and objective evaluation and feedback direction to correct pronunciation errors to improve oral English learning efficiency.