Recently, the international speech recognition big cow, former professor of Jonhs Hopkins University and Daniel Povey, the father of Kaldi, the voice recognition open-source tool, announced on Twitter that he will sign a working agreement with Xiaomi, based in Beijing.
In addition, Jonhs Hopkins also said that he will travel to Beijing before the end of 2019, and will recruit a small team to create a new generation of “PyTorch-y” Kaldi.
Today, Xiaomi President Lin Bin also announced the news on Weibo: the founder of the speech recognition open-source tool Kaldi, Daniel Povey, the voice and AI field, decided to join Xiaomi. He will develop the next generation of Kaldi in the Xiaomi group, and interested students will contact @Cui Baoqiu as soon as possible.
According to the data, Danial Povey is an expert in speech recognition. Daniel Povey has studied undergraduate and Ph.D. in Cambridge, England, and received his Ph.D. from the University of Cambridge in 2003.
He led the development of the speech recognition tool library Kaldi, which supports the training and prediction of a variety of speech recognition models. Many domestic and foreign speech technology companies’ R&D tests started from Kaldi. It integrates a variety of speech recognition models, including Hidden Markov and the latest deep learning neural network, which has an unshakable position in speech recognition.
It is speculated that Danial Povey will join Xiaomi this time and will greatly enhance Xiaomi’s future strength in voice and AI.