EUROSPEECH 2003 - INTERSPEECH 2003
One of the best challenges in Lip Reading is to apply this technology to an embedded device. In the current solutions the high use of resources, especially in reference to visual processing, makes the implementation in a small device very difficult. In this article a new, efficient and straightforward algorithm for detection and tracking of lips is presented. Lip Finding and Tracking is the first step in visual processing for Lip Reading. In our approach the Lip Finding is performed between a small amount of blobs, which should fulfill a geometric restriction. In terms of computational power and memory the proposed algorithm meets the requirements of an embedded device; on average less than 4 MHz^1 of CPU is required. This algorithm shows promising results in a realistic environment accomplishing successful lip finding and tracking in 94.2% of more than 4900 image frames.
Bibliographic reference. Perez, Jesus F. Guitarte / Lukas, Klaus / Frangi, Alejandro F. (2003): "Low resource lip finding and tracking algorithm for embedded devices", In EUROSPEECH-2003, 2253-2256.