Coupled Dictionaries for Exemplar-Based Speech Enhancement and Automatic Speech Recognition

Baby, Deepak; Virtanen, Tuomas; Gemmeke, Jort F.; hamme, Hugo Van

doi:10.1109/taslp.2015.2450491

Cited by 33 publications

(19 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Using this approach, modulation spectrogram features are being presented for the exemplar-based tasks [11].…”

Section: Related Workmentioning

confidence: 99%

MFCC using Speech Recognition in Computer Applications for Deaf

2019

IJRTE

View full text Add to dashboard Cite

Speech Recognition has lately evolved as a beneficial computer technology involving various interactive speech based applications in this domain. For communicating with one another, speech forms the basic foundation. Utilizing this way of communication in the technology domain, the technique of speech recognition is formulated. This technique considers the input speech in order to extract significant information thereby drawing accurate decision related to the concerned text. Here the research proposes a technique by effectively incorporating speech recognition approach for handling web portal via voice (be it from any place or time) for accessing data. Advent and growth in technology has led to formation of STT (speech–to–text) conversion model which generates a text format of the speech that’s beneficial for the Deaf individuals and in other realms too. Data mining has achieved great success in examining the acoustic features related to sound and speech. The existing research employs the MFCC (Mel Frequency Cepstrum Coefficients) for extracting acoustic features of a voice that is identifying the gender of a given voice. Following are the stages in the system proposed: 1. speech preprocessing, 2. feature extraction, 3. converting to text4. Classification5. Displaying the web portal and 6. Storing the web portal. The research depicts speech recognition and analysis as a speech process. The STT (speech to text) model includes two levels, first being the document level and second is the sentence level. The proposed system is beneficial for the deaf individuals for accessing the data be it from any place or time and in other realms too. The system yields in great accuracy in the given time span.

show abstract

“…Using this approach, modulation spectrogram features are being presented for the exemplar-based tasks [11].…”

Section: Related Workmentioning

confidence: 99%

MFCC using Speech Recognition in Computer Applications for Deaf

2019

IJRTE

View full text Add to dashboard Cite

show abstract

“…The earliest work on NMF based dereverberation [3] uses a convolutive NMF (referred as C-NMF) model for the reverb spectrogram. Since then many modifications to this have been proposed both in single-channel [4,5,6,7,8] and multi-channel scenario [9].…”

Section: Introductionmentioning

confidence: 99%

“…The C-NMF model for speech dereverberation was improved by additionally incorporating a NMF model for clean speech [5,6]. Various supervised approaches to handle reverberation in noisy environments have also been proposed [7,10,11]. Different regularization on RIRs in singlechannel [11,12] and multi-channel [13] scenario have been proposed leading to better speech enhancement.…”

Section: Introductionmentioning

confidence: 99%

A Non-convolutive NMF Model for Speech Dereverberation

2018

View full text Add to dashboard Cite

Reverberation corrupts speech recorded using distant microphones, resulting in poor speech intelligibility. We propose a single-channel, supervised non-negative matrix factorization (NMF) based dereverberation method, in contrast to the convolutive NMF (CNMF) based methods in literature. Recent supervised approaches use a CNMF model for reverberation and a NMF model for clean speech spectrogram to obtain enhanced speech by directly estimating the clean speech activations. In the proposed method, with a separability assumption on the room impulse response (RIR) spectrogram, the reverb speech can be decomposed into bases and activations using conventional NMF. Using these reverb activations, the clean speech activations are estimated to obtain enhanced speech. The proposed model (i) helps in imposing meaningful constraints on the RIR in both frequency-and time-domains to achieve improved enhancement (ii) leads to a framework that can include a NMF model for noise. (iii) gives a better interpretation of the effects of reverberation in the NMF context. We evaluate and compare the enhancement performance of the algorithm on reverb and noisy conditions, simulated using TIMIT utterances and REVERB challenge RIRs. The proposed method performs better than existing C-NMF based methods in objective measures, such as cepstral distance (CD) and speech-toreverberation modulation energy ratio (SRMR).

show abstract

“…In particular, speech recognition in exemplar based approaches is performed either using the atom activations of the estimated sparse feature vector [3], [4], or using the minimum reconstruction error [5] between the test exemplar and its estimate. On the contrary, in feature based approaches, either the derived sparse vector [1] or the estimate of speech is used as a feature [6] for acoustic modeling.…”

Section: Introductionmentioning

confidence: 99%

“…On the contrary, in feature based approaches, either the derived sparse vector [1] or the estimate of speech is used as a feature [6] for acoustic modeling. For computing the sparse feature vector, approaches in [3], [4] use a single overcomplete dictionary while [5] use multiple dictionaries corresponding to different speech units. A gradient descent approach is used to learn a single overcomplete dictionary using the spectro-temporal representation in [1], while mel frequency cepstral coefficients (MFCC) of training speech data (frames) are used to obtain dictionary atoms in [6] and [2].…”

Section: Introductionmentioning

confidence: 99%

Class specific GMM based sparse feature for speech units classification

Sharma

Abrol

Dileep

et al. 2017

2017 25th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

Abstract-In this paper, features based on the sparse representation (SR) are proposed for the classification of speech units. The proposed method employs multiple dictionaries to effectively model variations present in the speech signal. Here, a Gaussian mixture model (GMM) is built using spectral features corresponding to frames of all the examples of a speech class. Multiple dictionaries corresponding to different mixture are learned using the respective speech frames. Given a train/test speech frame, minimum spectral distance measure from the GMM means is employed to select an appropriate dictionary. The selected dictionary is used to obtain the sparse feature representation, which is used for the classification of speech units. The effectiveness of the proposed feature is demonstrated using continuous density hidden Markov model (CDHMM) based classifiers for (i) classification of isolated utterances of E-set of English alphabet, (ii) classification of consonant-vowel (CV) segments in Hindi language and (iii) classification of phoneme from TIMIT phonetic corpus. Experimental results reveal that the proposed features outperforms existing feature representations for various speech units classification tasks.

show abstract

Coupled Dictionaries for Exemplar-Based Speech Enhancement and Automatic Speech Recognition

Cited by 33 publications

References 39 publications

MFCC using Speech Recognition in Computer Applications for Deaf

MFCC using Speech Recognition in Computer Applications for Deaf

A Non-convolutive NMF Model for Speech Dereverberation

Class specific GMM based sparse feature for speech units classification

Contact Info

Product

Resources

About