446 Hits in 7.1 sec

Speech and Language Recognition using MFCC and DELTA-MFCC

Samiksha Sharma, Anupam Shukla, Pankaj Mishra
2014 International Journal of Engineering Trends and Technoloy  
To train this model resilient back propagation algorithm and radial basis function neural network used and results are compared.  ...  A multi speaker Speech recognition and language recognizer proposed for these four Indian languages  ...  Experiment conducted on three language English-French-Mandarin data.  ... 
doi:10.14445/22315381/ijett-v12p286 fatcat:wpvuor3e5vdf7k2sjzn5ba5s5m

Structuring Broadcast Audio for Information Access

Jean-Luc Gauvain, Lori Lamel
2003 EURASIP Journal on Advances in Signal Processing  
Audio indexing must take into account the specificities of audio data such as needing to deal with the continuous data stream and an imperfect word transcription.  ...  One rapidly expanding application area for state-of-the-art speech recognition technology is the automatic processing of broadcast audiovisual data for information access.  ...  ACKNOWLEDGMENTS This work has been partially financed by the European Commission and the French Ministry of Defense.  ... 
doi:10.1155/s1110865703211033 fatcat:uouvssghczfrvbzekvj5chk32y

An RNN-based preclassification method for fast continuous Mandarin speech recognition

Sin-Horng Chen, Yuan-Fu Liao, Song-Mao Chiang, Saga Chang
1998 IEEE Transactions on Speech and Audio Processing  
A novel recurrent neural network-based (RNN-based) frontend preclassification scheme for fast continuous Mandarin speech recognition is proposed in this paper.  ...  Efficiency of the proposed scheme was examined by simulations in which we incorporate it with a hidden Markov model-based (HMM-based) continuous 411 Mandarin base-syllables recognizer.  ...  EXPERIMENTAL RESULTS Efficiency of the proposed method was examined by simulations using a continuous Mandarin speech data base uttered by a single male speaker.  ... 
doi:10.1109/89.650315 fatcat:okmw7hnxj5fxtilkbia3clhn2a

Design and Applications of Embedded Systems for Speech Processing [chapter]

Jhing-Fa Wang, Po-Chun Lin, Bo-Wei Che
2012 Embedded Systems - High Performance Systems, Applications and Projects  
Speech, and Signal Processing, pp. 593-596, Salt Lake City, Utah, USA, May 07-11, 2001. Itoh, Y. & Tanaka, K. (2002)  ...  The operation of the proposed system based on SMO involves a training phase and an identification phase.  ...  Data set Data set A Data set B Data set C Phase Parameter setting phase Evaluation phase Number of database sentence 50 Mandarin 50 Mandarin 50 Mandarin + 50 English Number of query sentence  ... 
doi:10.5772/38558 fatcat:gycpgnugwzhwbiio6mp4bnjqcq

Pitch-based gender identification with two-stage classification

Yakun Hu, Dapeng Wu, Antonio Nucci
2011 Security and Communication Networks  
rule on a scalar (i.e., pitch) is used.  ...  In this paper, we address the speech-based gender identification problem. Mel-Frequency Cepstral Coefficients (MFCC) of voice samples are typically used as the features for gender identification.  ...  Both GMMs of male speakers and female speakers are trained by Expectation Maximization (EM) algorithm, using the pitch feature vectors of all male speakers and all female speakers,respectively.  ... 
doi:10.1002/sec.308 fatcat:2y34g4flsrgqtictvyh723spwe

The effect of enhancing temporal periodicity cues on Cantonese tone recognition by cochlear implantees

Tan Lee, Shing Yu, Meng Yuan, Terence Ka Cheong Wong, Ying-Yee Kong
2014 International Journal of Audiology  
Test materials were Cantonese disyllabic words recorded from one male and one female speaker. Speech-shaped noise was added to clean speech.  ...  Electrical stimuli generated from the noisy speech with and without periodicity enhancement were presented via direct stimulation using a Laura 34 research processor.  ...  Speech stimuli used for training were recorded from one male and one female speaker.  ... 
doi:10.3109/14992027.2014.893374 pmid:24694089 pmcid:PMC4222519 fatcat:yiabqnfvtzfk7mybenkpoh4qeu

Personalize Mobile Access By Speaker Authentication [chapter]

Ke Chen
2002 Biometric Solutions  
On the one hand, we use text-dependent speaker verification for handset protection as the primary stage of our security system.  ...  On the other hand, a more sophisticated speaker authentication system consisting of text-independent speaker verification and verbal information verification is located in the authentication center of  ...  Li and T.Y. Wu for discussions and their help in simulations. The work described here was supported in part by an NSFC grant (60075017) and an MSR research grant on non-verbal speech analysis.  ... 
doi:10.1007/978-1-4615-1053-6_5 fatcat:jydnon5omndypf77okbqozmiue

Vocal Resonance

Rui Liu, Cory Cornelius, Reza Rawassizadeh, Ronald Peterson, David Kotz
2018 Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies  
We collected data from 29 subjects, demonstrate the feasibility of a prototype, and show that our DNN method achieved balanced accuracy 0.914 for identification and 0.961 for verification by using an LSTM-based  ...  We explore two machine-learning approaches that analyze voice samples from a small throat-mounted microphone and allow the device to determine whether (a) the speaker is indeed the expected person, and  ...  The views and conclusions contained in this document are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied, of the sponsors  ... 
doi:10.1145/3191751 fatcat:qcomnhdx55hffn7gw2ejuv5lsm

Designing an Intelligent Translation Software by Audio Processing Techniques

Neda Payande, Behname Ghavami
2016 Bulletin of Pure & Applied Sciences- Physics  
Some innovative methods were used for fast and automatic preparation of essential data to train neural network of generating tone for naming and sectioning of parts of speech.  ...  Systems that do not use training are called "speaker independent" systems. Systems that use training are called "speaker dependent".  ... 
doi:10.5958/2320-3218.2016.00006.3 fatcat:x7j7d2mhcrddnhph2vinbj77su

Language identification using acoustic log-likelihoods of syllable-like units

T. Nagarajan, H.A. Murthy
2006 Speech Communication  
Using these language-dependent syllable-like unit models, language identification is performed based on accumulated acoustic log-likelihoods.  ...  The training data of each of the languages is first segmented into syllablelike units and language-dependent syllable-like unit inventory is created.  ...  Even if the speech data used during training is limited, the n-gram statistics can very well be derived from the digital text and used for the language identification task.  ... 
doi:10.1016/j.specom.2005.12.003 fatcat:mzgkmuxearc2beu6r7sg7svuti

Imitation of second language sounds in relation to L2 perception and production

Yen-Chen Hao, Kenneth de Jong
2016 Journal of Phonetics  
More detailed predictions of the error patterns in the Imitation task based on Perception, Production, and Cascade models were compared.  ...  Experiment 1 targeted English speakers' learning of Mandarin tones, while Experiment 2 investigated Korean speakers' learning of English consonants.  ...  One female native speaker of Mandarin produced all the stimuli for the Identification and Imitation tasks, which were recorded at 44.1 kHz using the built-in microphone of a laptop (Compaq Presario CQ45  ... 
doi:10.1016/j.wocn.2015.10.003 fatcat:c5s4gvlhsjg23nxqizabkl3s3q

Bayesian classification for data from the same unknown class

Hung-Ju Huang, Chun-Nan Hsu
2002 IEEE Transactions on Systems Man and Cybernetics Part B (Cybernetics)  
training and classification.  ...  Our method, called homologous naive Bayes (HNB), is based on the naive Bayes classifier, a simple algorithm shown to be effective in many application domains.  ...  However, when the training data set is too small, this often yields and impedes the classification. To avoid this problem, another popular choice is for all .  ... 
doi:10.1109/3477.990870 pmid:18238113 fatcat:mk4pgcaqprdkrmoasvtlcxdq7a

Dialect Identification of Assamese Language using Spectral Features

Tanvira Ismail, L. Joyprakash Singh
2017 Indian Journal of Science and Technology  
As mentioned, we have developed the database and then Mel-Frequency Cepstral Coefficient has been used to extract the spectral features of the collected speech data.  ...  Findings: Research work done on dialect identification is relatively much less than that on language identification for which one of the reasons being dearth of sufficient database on dialects.  ...  English, based on clustering and supervised learning. 12 First the feature vectors using LP coefficients were obtained and clusters of vectors using the K-means algorithm were formed.  ... 
doi:10.17485/ijst/2017/v10i20/115033 fatcat:msdxlwjjsrfrzpxurcfxon4kdy

A spoken-access approach for chinese text and speech information retrieval

Lee-Feng Chien, Hsin-Min Wang, Bo-Ren Bai, Sun-Chien Lin
2000 Journal of the American Society for Information Science  
The encouraging results suggest that a Mandarin speech interface for information retrieval and digital library systems can, therefore, be developed.  ...  Based on utilization of the mono-syllabic structure of the Chinese language, the proposed approach can tolerate speech recognition errors by performing speech query recognition and approximate information  ...  Here, an algorithm based on a commonly-used word lexicon, a general-domain corpus and a key term selection strategy, is employed.  ... 
doi:10.1002/(sici)1097-4571(2000)51:4<313::aid-asi2>;2-i fatcat:dkwnorljk5bkjis6xhdfrad7sm

Improved learning algorithms for mixture of experts in multiclass classification

K. Chen, L. Xu, H. Chi
1999 Neural Networks  
However, it is reported in literature that the IRLS algorithm is of instability and the ME architecture trained by the EM algorithm, where IRLS algorithm is used in the inner loop, often produces the poor  ...  To tackle the expensive computation of the Hessian matrix and its inverse, we propose an approximation to the Newton-Raphson algorithm based on a so-called generalized Bernoulli density.  ...  This work was supported by the HK RGC Earmarked Grant CUHK 250/94E and National Science Foundation in china.  ... 
doi:10.1016/s0893-6080(99)00043-x pmid:12662629 fatcat:begzahm5zzc3njq6f4inveisu4
« Previous Showing results 1 — 15 out of 446 results