Is Baidu's speech recognition used in Iflytek?

Baidu speech recognition provides developers with high-quality and free voice services in the industry. Through the optimization of scene recognition, it provides voice solutions for car navigation, smart home and social chat industries, with an accuracy rate of over 90%, so that your application can draw "sound". Baidu is also doing voice technology, and now it has completely opened the voice recognition and speech synthesis technology for free, and individual developers can use it for free. Baidu brain team led by Andrew Ng is also providing support for speech recognition, speech synthesis and other technologies. In the future, Baidu's speech technology should not be underestimated. Iflytek is a pioneer in the domestic voice industry, and its applications range from military to civilian, including mobile Internet, education, car networking, smart home and so on. Iflytek provides voice cloud and supports many dialects. Most speech recognition functions in China use the interface of Iflytek, such as Gaode map. Internet companies such as Baidu, sogou and Tencent also have their own speech recognition technologies, which are developing rapidly relying on their rich user resources and data. As far as speech recognition technology is concerned, the above companies all use the Deep Neural Network (DNN) algorithm, and there may be some differences in noise reduction. More differences are found in the corpus. After years of accumulation, Iflytek has a rich corpus. Of course, the big data of those Internet companies should not be underestimated. From personal experience, Iflytek is the best dialect recognition at present, and punctuation marks are automatically given according to pause and semantics.