We present an algorithm using convolutive non-negative matrix factorization (CNMF) to create noise-robust features for automatic speech recognition (ASR).
ABSTRACT. We present an algorithm using convolutive non-negative matrix factorization (CNMF) to create noise-robust features for automatic.
Motivation. • MFCCs and Mel filterbank energies commonly used as acoustic features for automatic speech recognition (ASR). • These features are not robust�...
They showed that the NMF-based approach gives better ASR performance than log-mel features or denoising the speech. ... ... On a training set�...
We present an algorithm using convolutive non-negative matrix factorization (CNMF) to create noise-robust features for automatic speech recognition (ASR).
CNMF-Based Acoustic Features for Noise-robust ASR. Vaz, C., B Dimitriadis, D., Thomas, S., & Narayanan, S. S. In Proceedings of IEEE International�...
People also ask
What is the acoustic model and language model in ASR?
What is noise robust speech recognition?
Speech Processing: Robust Speech Recognition. Paper Title: CNMF-BASED ACOUSTIC FEATURES FOR NOISE-ROBUST ASR. Authors: Colin Vaz, University of Southern�...
First, an unsupervised technique using non-negative matrix factorization (NMF) discovers phone-sized time–frequency patches into which speech can be decomposed.
Jul 16, 2019 � We propose an algorithm to extract noise-robust acoustic fea- tures from noisy speech. We use Total Variability Modeling in.
Sep 1, 2011 � Using local acoustic context for learning noise bases. The success of CNMF-based approaches depends on learning reliable bases for the�...