论文部分内容阅读
语音识别作为人机交互的重要接口,是推动人工智能发展的重要技术。在语音识别领域,深度学习技术能够使声学模型的识别性能得到显著性提升,已然成为新的研究热点。文章从深度学习在语音识别领域取得的显著成就出发,介绍了语音识别的基本原理和系统组成,详细阐述了深度神经网络在语音识别声学建模当中的应用,最后探讨了深度学习在语音识别领域的重要发展方向。
As an important interface of human-computer interaction, speech recognition is an important technology to promote the development of artificial intelligence. In the field of speech recognition, deep learning technology can significantly improve the recognition performance of acoustic model, which has become a new research hotspot. Based on the deep achievements made in the field of speech recognition, the article introduces the basic principles and system components of speech recognition, elaborates the application of deep neural networks in the speech recognition and acoustic modeling, and finally discusses the application of deep learning in the field of speech recognition The important direction of development.