报告题目: Silent Speech Interfaces
报告人: Prof. Ian Mcloughlin
Dept. of Electronic Engineering & Information Science, School of Information Science and Technology, the University of Science and Technology of China
报告时间：4月21日 星期一 下午1:30
报告地点：浙江大学玉泉校区控制系智能系统与控制研究所楼304会议室

报告摘要：Silent Speech Interfaces (SSI) are quiet or inaudible methods of vocal human-computer communications. They include many types of technology and basic approaches, including whisper input, video lip-reading, Doppler detection, and even nerve and brain sensing interfaces. Application domains include ASR, authentication, medical prosthesis and mobile communications. In fact, any speech-based system which needs to operate silently (or nearly silently) at times, will require this technology. One of the most important future applications will be future ubiquitous computing interaction and wearable devices such Google's Glass, which use speech for human-computer interfacing.
Imagine if you need to speak a password to a wearable computer, or discuss something private with it, but you are sat on the train or in a meeting. You probably don't want other people nearby to overhear what you are saying. That is when you need a silent speech interface.
In this talk, Professor McLoughlin will first give an overview of current SSI methods before focussing on practical research involving two types of SSI: whisper-to-speech reconstruction, and inaudible vocal tract mapping. The whisper-to-speech reconstruction technique uses machine-learning techniques to reconstruct nearly-natural speech from whispers. The inaudible vocal tract mapping system is a novel method that uses a convenient gap in the frequency spectrum, just above the threshold of human hearing. He will describe both methods and demonstrate how they can be used in smartphones and other speech devices.

报告人简介：Professor Ian Mcloughlin is currently working at the Dept. of Electronic Engineering & Information Science, the University of Science and Technology of China. Professor Ian Mcloughlin has worked >10 years in industry and >10 years in academia. He has held every industrial engineering job grade from research student through to principal engineering (on three continents �C for government, multinational, and SME). He also held every academic professional job grade. His research focus covers Embedded systems & wireless, Computer Architecture, Space technology, Green and sustainable systems, Speech and audio and Reverse engineering.