Update:2018.08.22
British magazine "Impact" interviewed Prof. Takeshi Saitoh and introduces his research on multi-modal silent speech recognition technology in its recent issue (Volume 2018, Number 3, June 2018, pp. 47-49(3)).
Lip reading (or visual speech recognition) is a technology to estimate the speech content from only video data without using audio data, and it is expected to be applied as the communication support system for disabled people and bimodal speech recognition.
In the article, Silent Speech Recognition Workshop (SSRW) which is mainly organized by Prof. Saitoh is also introduced.
You can read the article from HERE.