Audio transcribing device tells voices apart from speech
A computer programmer has invented a digital device that is said to be able to distinguish different voices from a conversation and automatically transcribe words into a script, reports the Jiangsu-based Yangtze Evening News.
It's said to be the first voice transcribing device in China which can recognize different people engaged in a real-time conversation.
Deng Chong with his audio transcribing device [Photo: sohu.com]
Deng Chong from Alibaba's cloud computing service Aliyun spent 6 months creating the program, which consists of a microphone which feeds audio into a specialized chip. Using Aliyun's computing method, the device is said to be able to achieve 90% accuracy in voice recognition.
Deng says he was inspired after a fight with his girlfriend, saying he wished he had the ability to review who said what during their argument.
Aliyun says its now setting up a new team to work on the mass production of the device.
Aliyun says its initial target for sales will be media outlets to help with interview transcriptions.