In this paper, we introduce a novel vector quantization (VQ) scheme for distributing the quantization error equally among the quantized dimensions. Afterwards, the proposed VQ sch...
A key step in the process of lip-reading is determining the shape of the speaker’s lips. This has previously been achieved through an energy method known as “snakes”, howeve...
In this paper, we propose to use the scaling ambiguity of convolutive blind source separation for shortening the unmixing filters. An often used approach for separating convoluti...
In this paper, we re-examine the recently proposed distributed state estimators based on quantized innovations. It is widely believed that the error covariance of the Quantized In...
We address multistream sign language recognition and focus on efficient multistream integration schemes. Alternative approaches are investigated and the application of Product-HM...