We propose Cross-Channel Spectral Subtraction (CCSS), a source separation method for recognizing meeting speech where one microphone is prepared for each speaker. The method quick...
The importance of faces in human interaction explains the desire for synthetic faces as a communication vehicle in computer graphics. Unfortunately, animating a face is a very com...
This paper presents a novel approach for recognizing and interpreting dimensions in engineering drawings. It starts by detecting potential dimension frames, each comprising only t...
A method for segmenting and recognizing text embedded in video and images is proposed in this paper. In the method, multiple segmentation of the same text region is performed, thu...
This paper proposes a new method for personal identification based on iris recognition. The method consists of three major components: image preprocessing, feature extraction and ...