Patent speech-to-text innovation:
improve accuracy and dynamic biometric authentication

'Your identity is talking'​

What is Luvius?

Spoken words and Lip movement in sync
shows over-articulation:

Spoken words

Sound signal

The characteristic of a phoneme is stored in the letter collection database and can be used for recognizing corresponding moments from speech

Lip movement

Light reflection sensor

The mouth position reveals the following phoneme (letter sound)”

The Product

Hardware

Headset

(Phonebox device and Smartphone not yet developed)

Software

Object oriented language, C++, application<1mb

Demo & technical data
(real-time spoken-recorded-prepared-analysed
and shown on screen)

Growth S2T

Devices & Application

Authentication

Innovate the speech-to-text market, disrupt the authentication market

What is unique about Luvius?

Existing technology
  • Solutions for speech recognition
  • Sound & context analyses
  • Camera images and sound
  • Doppler radar
  • Low frequency waves
  • WiFi-signal
Innovation technology
  • LR technology
  • Lip reading with light reflection
  • Speech and voice recognition.

Luvius lip reading with light reflection

Improvements of known speech recognition solutions:

1) faster: the phoneme is known when sound starts

2) Specific: no doubt which phonemes are out of reach

3) Simple: ownly phoneme characterics are compared with
corresponding items in de database

4) Less energy: no complex queries or logarithms are needed

Improvement of known product:

1) Context of what is spoken is not relevant (collection of spoken words in the past is not needed)

2) Noisy environment is ignored (light reflection is leading)

3) Random generated words for authentication will do (no registrated passphrase needed)

4) Authentication without biometric data (just collecting indirect light reflection of mouth movements)

Application fields

Medical services

Building access

Smart payments

Aviation

Contact

Close Menu