隐马尔可夫模型（ＨＭＭ）技术是语音识别中应用较为成功的算法，但它的缺点影响了其精度、速度、硬件实现和推广应用，神经网络（ＮＮ）具有并行性、强的分类能力和易于硬件实现等优点。将ＮＮ与ＨＭＭ相结合构成混合网络，能克服ＨＭＭ与ＮＮ的缺点，保留双方的优点，本文详细评述了目前在语音识别中应用的由ＨＭＭ和ＮＮ构成的四种混合网络。通过对其结构、识别性能和特点的分析，可以看出ＨＭＭ和ＮＮ构成的混合网的性能明显优于纯ＨＭＭ和ＮＮ，是更适于语音识别的网络。 Hidden Markov Model (HMM) is a more successful algorithm in speech recognition, but its shortcomings affect its accuracy, speed, hardware implementation and application. Neural Network (NN) has the advantages of parallelism, strong classification ability and Easy to implement hardware and so on. Combining NN and HMM to form a hybrid network can overcome the shortcomings of both HMM and NN and retain the advantages of both. In this paper, four hybrid networks composed of HMM and NN, which are currently used in speech recognition, are reviewed in detail. Through the analysis of its structure, recognition performance and characteristics, it can be seen that the performance of hybrid network composed of HMM and NN is obviously better than that of pure HMM and NN, which is more suitable for speech recognition network.