Exploring the Teaching Mode of English Audiovisual Speaking in Multimedia Network Environment

Introducing multimedia network tools in English audiovisual teaching and building a new model of network-based can make more line with students’ cognitive thinking characteristics processes. This improve the overall efficiency schools. Computers have been widely used language evaluation speech recognition for learning, technology is an important reflection level learning. The large amount signal data, complex pronunciation changes, high dimensionality feature parameters learning process it difficult to identify features. computational volume too large, which requires hardware resources software realize high-speed processing massive signals. To address problem low rate pronunciation, this study proposes sound algorithm based on adaptive particle swarm optimization (PSO) matching pursuit (MP) sparse decomposition. firstly improves parameter setting PSO population evolution rate, establishes PSO, realizes optimized MP continuous Gabor super-complete atomic set constructed space search property optimal evolutionary process. Finally, realized by support vector machine (SVM) algorithm. test results show that misjudgement different mispronunciations less than 1% when system evaluate level. It proves method effectively detect mispronunciation has accuracy.