Deciphering speech from lip movements
Read about Liopa's SRAVI application in CNET from the CES2021 show
01
“I see Liopa’s technology as fundamental to the recovery and wellbeing of critically ill patients.”
-Dr Shondipon Laha, Lancashire Teaching Hospitals NHS Trust
Our AI-based LipRead technology is currently being applied in three areas
SRAVI: Bringing a voice to the voiceless
Our smartphone-based lipreading app, SRAVI, is a communications aid for those who cannot speak. It’s being trialled in UK NHS hospitals with patients in ICU.
Patients with tracheostomies, laryngectomies and/or prolonged intubation have great difficulty speaking. SRAVI helps these patients convey important information quickly. SRAVI can also be used at home and to help with a range of speech debilitating conditions.
Keyword spotting in silent video, e.g. CCTV
Our LipRead technology can detect, in silent video, the occurrence of specified words or phrases.
LipRead can identify whether certain things were said, when an audio track is unavailable. We’re currently researching this in a DASA project involving silent CCTV footage that could assist law enforcement to use video footage more effectively in fighting crime.
Research and development apps
We’re constantly reviewing how Liopa’s flagship LipRead technology can be applied in different ways, including:
Silent communicator – an app that understands your lip syncing in gaming or video calling
Active speaker detection – in situations where more than one speaker is talking, e.g. in-car systems – our technology picks out the main speaker
Cloud services – for use of lipreading in video
Find out more – Research Apps