Mouth to Speech

The Mouthesizer tracks the shapes made by the mouth and uses it to control a music synthesizer. It's an interesting idea to segment the region between the lips and to use the width and height of the segmented region to control the characteristics of the generated sound. If lip movements can be mapped to actual human speech, this could be used by visually impaired people to generate speech from the lip movements of people in front of them.. (the opposite of this)

Tags: mouthesizer, speech, image processing, science, signal processing