Automatic speech recognition (ASR) systems, trained on speech signals from close-talking microphones, generally fail in recognizing far-field speech. In this paper, we present a Hilbert Envelope based feature extraction technique to alleviate the artifacts introduced by room reverberations. The proposed technique is based on modeling temporal envelopes of the speech signal in narrow sub-bands u...

This paper is focused on improving visual Czech speech synthesis. Our aim was the design of a highly natural and realistic talking head with a realistic 3D face model, improved co-articulation, and a realistic model of inner articulatory organs (teeth, the tongue and the palate). Besides very good articulation our aim was also expression of the mimic and emotions of the talking head. The intell...

We propose a blind dereverberation method based on spectral subtraction using a multi-channel least mean squares (MCLMS) algorithm for distant-talking speech recognition. In a distant-talking environment, the channel impulse response is longer than the short-term spectral analysis window. By treating the late reverberation as additive noise, a noise reduction technique based on spectral subtrac...