Intelligent speech recognition and voice response have become important technologies in the era of mobile Internet. Apple, Microsoft, Google (microblogging), Amazon and other giants have layouts. In the accuracy of human speech recognition, various manufacturers have also launched an arms race. The latest news shows that the Microsoft R&D team achieved a human-acceptance accuracy in recognition rate. Microsoft's artificial intelligence R&D team recently published a paper on speech recognition, which mentions the recognition technology they developed, which has achieved a vocabulary error rate of 5.9%. Foreign media pointed out that such an error rate or accuracy rate has been basically close to that of normal humans. Xuedong Huang, the head of Microsoft's speech recognition research, said: "We have achieved the same level of recognition as humans. This is a historic breakthrough." It is reported that the teams of various manufacturers are doing everything possible to improve the accuracy of speech recognition and reduce the vocabulary error rate. In the last month, the latest score of the error rate has dropped to 6.3%, and Microsoft's latest paper has further achieved breakthroughs. . Unlike other innovative research that may be shelved, speech recognition technology can be quickly applied to actual products, so this research will also have a significant impact on Microsoft's actual business, such as Microsoft's voice assistant Cortana (Chinese commonly known as "small" Na", Xbox game platform, Windows operating system deploys the latest speech recognition technology. In the field of intelligent speech recognition, Microsoft's Cortana came to the fore, and the industry evaluation has surpassed Apple's Siri. At present, Microsoft has covered the voice assistants with various computing tools, such as mobile phones, tablets, and computer operating systems. In order to improve the accuracy of speech recognition, Microsoft researchers used a neural network to store a large amount of data, which can help the recognition software to more intelligently identify patterns of human speech expression. In addition, Microsoft also needs to use voice information and picture information to enable neural networks to more efficiently use the massive amounts of data stored. However, in the eyes of researchers, reaching the level of humanity is not the ultimate goal of research, and the current identification technology is not perfect. The Microsoft team said that it will further improve the accuracy of speech recognition. In addition, in order to ensure more practical applications of recognition technology, Microsoft plans to achieve high-precision recognition in a more noisy environment, such as in a lively restaurant, crowded streets, or in a stormy environment, Microsoft I hope that in an abnormal environment, the voice system can also accurately understand what people are saying. It is also reported that in the era of smart phones, the status of speech recognition technology is more important than in the past, and the current poor performance of Apple Siri has triggered the attention of Apple executives. Just recently, Apple has hired a professor of artificial intelligence technology from Carnegie Mellon University in the United States, and will be responsible for the formation of Microsoft's artificial intelligence research team. One of the missions of this team is to greatly improve Siri's recognition of speech and intelligent answering questions. "IQ." Waterproof Speaker,Full Range Loud Speaker,Entrance System Speaker,Waterproof Multimedia Speaker Jiangsu Huawha Electronices Co.,Ltd , https://www.hnbuzzer.com