Improving Robustness of Codebook-Based Noise Estimation Approaches With Delta Codebooks

Rosenkranz, Tobias; Puder, Henning

doi:10.1109/tasl.2011.2172943

Cited by 17 publications

(8 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [4], an efficient codebook search method is introduced to reduce the speech codebook search range for the ML estimate. In [5], the MMSE estimate requires a relatively small speech codebook with only 32 entries without severe degradation of the performance. In our experiment, the NB and WB speech LSF codebooks are trained using the LBG [6] method with 10-min of speech from the TIMIT database using the Itakura-Saito measure [2].…”

Section: Resultsmentioning

confidence: 99%

Blind bandwidth extension for codebook-based Bayesian speech enhancement

Kim

Kang

2014

The 18th IEEE International Symposium on Consumer Electronics (ISCE 2014)

View full text Add to dashboard Cite

This paper proposes a blind bandwidth extension (BBE) technique that improves the quality of narrowband (NB) speech signal enhanced by the codebook-based Bayesian speech enhancement method with minimum mean square error (MMSE) estimate. The MMSE weight vector of the NB speech envelope codebook is used to estimate the wideband (WB) speech envelope. An adaptive double shifting method is introduced in the BBE to obtain an enhanced WB excitation signal. Its performance is assessed via the log-spectral distortion (SD) and informal listening tests.

show abstract

Section: Resultsmentioning

confidence: 99%

Blind bandwidth extension for codebook-based Bayesian speech enhancement

Kim

Kang

2014

The 18th IEEE International Symposium on Consumer Electronics (ISCE 2014)

View full text Add to dashboard Cite

show abstract

“…In (19), denotes the iteration index, is the index of the speech utterance in the database, denote the missing data of the EM algorithm, which are the sequence of the underlying states and speech gains. Furthermore, is the posterior state probability, which is defined as equation (21) at the bottom of the page.…”

Section: A Off-line Parameter Training Of Speech and Noise Sarhmmsmentioning

confidence: 99%

“…To address noise suppression in non-stationary environments, auto-regressive hidden Markov models (ARHMM) [15]- [17] and codebooks [18]- [21] have been used successfully to model the statistics of speech and noise for speech enhancement. In these methods, the speech and noise signals are modeled as AR processes [22].…”

Section: Introductionmentioning

confidence: 99%

Sparse Hidden Markov Models for Speech Enhancement in Non-Stationary Noise Environments

Feng

Bao

Kleijn

2015

IEEE/ACM Trans. Audio Speech Lang. Process.

View full text Add to dashboard Cite

We propose a sparse hidden Markov model (HMM)based single-channel speech enhancement method that models the speech and noise gains accurately in non-stationary noise environments. Autoregressive models are employed to describe the speech and noise in a unified framework and the speech and noise gains are modeled as random processes with memory. The likelihood criterion for finding the model parameters is augmented with an regularization term resulting in a sparse autoregressive HMM (SARHMM) system that encourages sparsity in the speech-and noise-modeling. In the SARHMM only a small number of HMM states contribute significantly to the model of each particular observed speech segment. As it eliminates ambiguity between noise and speech spectra, the sparsity of speech and noise modeling helps to improve the tracking of the changes of both spectral shapes and power levels of non-stationary noise. Using the modeled speech and noise SARHMMs, we first construct a noise estimator to estimate the noise power spectrum. Then, a Bayesian speech estimator is derived to obtain the enhanced speech signal. The subjective and objective test results indicate that the proposed speech enhancement scheme can achieve a larger segmental SNR improvement, a lower log-spectral distortion and a better speech quality in stationary noise conditions than state-of-the-art reference methods. The advantage of the new method is largest for non-stationary noise conditions. Index Terms-Gain modeling, non-stationary noise, sparse autoregressive hidden Markov model (ARHMM), speech enhancement. 2329-9290

show abstract

“…Speech enhancement algorithms which employ trained models, such as codebooks [24][25][26][27][28], hidden Markov models (HMM) [29][30][31], Gaussian mixture models (GMM) [32], non-negative matrix factorization (NMF) models [33], dictionaries [34], etc., for speech and noise data are able to process noisy speech with sufficient accuracy even under nonstationary noise conditions. For example, codebook-based speech enhancement (CBSE) algorithms [25,26] estimate the noise power spectrum for short segments of noisy speech, thus tracking nonstationary noise better than the buffer-based methods [18].…”

Section: Introductionmentioning

confidence: 99%

Robust Bayesian estimation for context-based speech enhancement

Naidu

Srinivasan

2014

J AUDIO SPEECH MUSIC PROC.

View full text Add to dashboard Cite

Model-based speech enhancement algorithms that employ trained models, such as codebooks, hidden Markov models, Gaussian mixture models, etc., containing representations of speech such as linear predictive coefficients, mel-frequency cepstrum coefficients, etc., have been found to be successful in enhancing noisy speech corrupted by nonstationary noise. However, these models are typically trained on speech data from multiple speakers under controlled acoustic conditions. In this paper, we introduce the notion of context-dependent models that are trained on speech data with one or more aspects of context, such as speaker, acoustic environment, speaking style, etc. In scenarios where the modeled and observed contexts match, context-dependent models can be expected to result in better performance, whereas context-independent models are preferred otherwise. In this paper, we present a Bayesian framework that automatically provides the benefits of both models under varying contexts. As several aspects of the context remain constant over an extended period during usage, a memory-based approach that exploits information from past data is employed. We use a codebook-based speech enhancement technique that employs trained models of speech and noise linear predictive coefficients as an example model-based approach. Using speaker, acoustic environment, and speaking style as aspects of context, we demonstrate the robustness of the proposed framework for different context scenarios, input signal-to-noise ratios, and number of contexts modeled.

show abstract

Improving Robustness of Codebook-Based Noise Estimation Approaches With Delta Codebooks

Cited by 17 publications

References 25 publications

Blind bandwidth extension for codebook-based Bayesian speech enhancement

Blind bandwidth extension for codebook-based Bayesian speech enhancement

Sparse Hidden Markov Models for Speech Enhancement in Non-Stationary Noise Environments

Robust Bayesian estimation for context-based speech enhancement

Contact Info

Product

Resources

About