Reverberation Modeling for Robust Speech Recognition
نویسندگان
چکیده
The REMOS (REverberation MOdeling for Speech recognition) concept for reverberation-robust distanttalking speech recognition [1] is presented in this paper. REMOS extends a conventional hidden Markov model (HMM) trained on close-talking data with a reverberation model describing the acoustical environment. The combination of both models is performed during recognition to match the reverberant observation. Since varying acoustic conditions only require a reestimation of the reverberation model, REMOS is significantly more flexible than recognition systems trained on reverberant data.
منابع مشابه
Combating reverberation in large vocabulary continuous speech recognition
Reverberation leads to high word error rates (WERs) for automatic speech recognition (ASR) systems. This work presents robust acoustic features motivated by subspace modeling and human speech perception for use in large vocabulary continuous speech recognition (LVCSR). We explore different acoustic modeling strategies and language modeling techniques, and demonstrate that robust features with a...
متن کاملA Simplified Decoding Method for a Robust Distant-talking Asr Concept Based on Feature-domain Dereverberation
A simplified decoding method for the concept of REverberation MOdeling for Speech recognition (REMOS) [1] is proposed. In order to achieve robust distant-talking Automatic Speech Recognition (ASR), the REMOS concept uses a combination of clean-speech HMMs and a reverberation model to perform feature-domain dereverberation during decoding. The simplified decoding/dereverberation method proposed ...
متن کاملA Combined Approach for Estimating a Feature-domain Reverberation Model in Non-diffuse Environments
A combined approach for estimating a feature-domain reverberation model suitable for the robust distant-talking automatic speech recognition concept REMOS (REverberation MOdeling for Speech recognition) [1] is proposed. Based on a few calibration utterances recorded in the target environment, the combined approach employs ML estimation and blind estimation of the reverberation time to determine...
متن کاملAn Information-Theoretic Discussion of Convolutional Bottleneck Features for Robust Speech Recognition
Convolutional Neural Networks (CNNs) have been shown their performance in speech recognition systems for extracting features, and also acoustic modeling. In addition, CNNs have been used for robust speech recognition and competitive results have been reported. Convolutive Bottleneck Network (CBN) is a kind of CNNs which has a bottleneck layer among its fully connected layers. The bottleneck fea...
متن کاملRobust Features and System Fusion for Reverberation-robust Speech Recognition
Reverberation in speech degrades the performance of speech recognition systems, leading to higher word error rates. Human listeners can often ignore reverberation, indicating that the auditory system somehow compensates for reverberation degradations. In this work, we present robust acoustic features motivated by the knowledge gained from human speech perception and production, and we demonstra...
متن کامل