The experiment shows the method for audio classification based on 1-state HMM could decrease the misrecognition effectively and increase the accuracy of audio classification.

• 实验结果表明基于单状态的HMM模型音频分类方法能有效地减少误识率，提高音频分类的精确度。