Help
RSS
API
Feed
Maltego
Contact
Domain > www.bavieca.org
×
Welcome!
Right click nodes and scroll the mouse to navigate the graph.
×
More information on this domain is in
AlienVault OTX
Is this malicious?
Yes
No
DNS Resolutions
Date
IP Address
2015-03-10
216.34.181.97
(
ClassC
)
2024-08-28
204.68.111.100
(
ClassC
)
Port 80
HTTP/1.1 200 OKServer: nginxDate: Wed, 28 Aug 2024 09:32:46 GMTContent-Type: text/htmlContent-Length: 16898Connection: keep-alivevary: Accept-Encodingvary: Hostlast-modified: Sun, 16 Jun 2013 17:36:05 GMTetag: 4202-4df48ea192b40accept-ranges: bytescache-control: max-age3600expires: Wed, 28 Aug 2024 10:32:30 GMTx-from: sfp-ioweb82-1vary: Accept-Encoding !DOCTYPE html PUBLIC -//W3C//DTD XHTML 1.0 Transitional//EN http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd>!-- Bavieca Speech Recognition Toolkit Webpage -->html xmlnshttp://www.w3.org/1999/xhtml>head>meta http-equivContent-Type contenttext/html; charsetutf-8 />title>Bavieca.org/title>link relshortcut icon href./images/favicon.ico.png>link hrefstyle.css relstylesheet typetext/css/>/head>body>!-- header -->div idheader>div idslogan>Speech Recognition /div>div idlogo>a href#>The Bavieca ASR toolkit/a>/div>div idmenu>ul>li classactive>a hrefindex.html>Home/a>/li>li>a hreftools.html>Tools and API/a>/li>li>a hrefusingBavieca.html>Using Bavieca/a>/li>li>a hrefjavaAPI.html>Java API/a>/li>li>a hrefresources.html>Resources/a>/li>li>a hrefdocumentation.html>Code documentation/a>/li>/ul>/div>!-- google analytics -->script typetext/javascript> var _gaq _gaq || ; _gaq.push(_setAccount, UA-36935847-1); _gaq.push(_setDomainName, bavieca.org); _gaq.push(_trackPageview); (function() { var ga document.createElement(script); ga.type text/javascript; ga.async true; ga.src (https: document.location.protocol ? https://ssl : http://www) + .google-analytics.com/ga.js; var s document.getElementsByTagName(script)0; s.parentNode.insertBefore(ga, s); })();/script>/div>!-- end header -->!-- main -->div idmain>script srchttp://code.jquery.com/jquery-latest.min.js>/script>div idsidebar>h3>a idabout href#>About/a>/h3>ul classaboutDrop>li>a idwelcomeItem href#>Welcome/a>/li>li>a idlicenseItem href#>License/a>/li>li>a idbenchmarksItem href#>Benchmarks/a>/li>li>a idpublicationsItem href#>Publications/a>/li>/ul>h3>a iddownload href#>Installation/a>/h3>ul classdownloadDrop>li>a iddownloadItem href#>Download/a>/li>li>a idinstallItem href#>Install/a>/li>/ul>script>$(a#about).click(function(){ $(ul.aboutDrop).slideToggle(fast); });/script>script>$(a#download).click(function(){ $(ul.downloadDrop).slideToggle(fast); });/script>script>$(a#installation).click(function(){ $(ul.installationDrop).slideToggle(fast); });/script>/div>!-- end sidebar -->div idwelcome classstd>h1>Welcome/h1>p>b>Bavieca/b> is an open-source speech recognition toolkit intended for speech research and as a platform for the development of speech-enabled solutions by non-speech experts. It supports common acoustic modeling and adaptation techniques based on continuous density hidden Markov models (CD-HMMs), including discriminative training. Bavieca includes two efficient decoders based on dynamic and static expansion of the search space that can operate in batch and live recognition modes. The Bavieca toolkit offers a simple and modular design with an emphasis on efficiency, scalability and reusability. Bavieca exhibits competitive results on a classbenchmarksShow href#>standard benchmarks/a> and has been successfully used on a number of research projects addressing both read and conversational childrens speech as well as conversational adults speech./p> p>Bavieca has been entirely developed by a hrefmailto:dani@bltek.com>b>Daniel Bolaños/b>/a> at a hrefhttp://www.bltek.com>Boulder Language Technologies/a> and it is currently used in several a classpublicationsShow href#>research projects/a> and products. Bavieca is currently at the Beta development stage and its development is still a work in progress. An article introducing the Bavieca speech recognition toolkit has been pusblished in the IEEE Spoken Language Technology Workshop 2012/p> a href./articles/baviecaII.pdf>D. Bolaños, “The Bavieca Open-Source Speech Recognition Toolkit”. i>In Proceedings of IEEE Workshop on Spoken Language Technology (SLT)/i>, December 2-5, 2012, Miami, FL./a>div classline-separator>/div>h2>Design/h2>p>The list below summarizes some of the main design principles of Bavieca/p>div idfeatureList>ul> li>Written entirely in C++ making extensive use of the Standard Template Library (STL)/li> li>Small code base, ≈ 30,000 lines of code and ≈ 100 C++ classes/li> li>Reduced set of command line tools (about 25) that serve specific purposes such as accumulating sufficient statistics or estimating model parameters according to some estimation criterion/li> li>Extensive reuse of code across tools/li> li>Application Programming Interface (API) that enables develoment of stand-alone applications that exploit Baviecas speech recognition capabilities/li> li>Linear Algebra support through BLAS and LAPACK/li>/ul>/div>div classline-separator>/div>h2>Features/h2>p>The list below summarizes the main features of the Bavieca toolkit./p>div idfeatureList> ul> li>Large vocabulary continuous speech recognition ul> li>Dynamic search decoder with support for cross-word triphone and pentaphone HMMs/li> li>Weighted Finite State Acceptor (WFSA) based speech decoder and efficient WFSA network builder (cross-word triphones)/li> li>Efficient computation of emission probabilities thanks to the use of the nearest neighbor approximation, partial distance elimination and support for Single Instruction Multiple Data (SIMD) parallel computation (x86 architecture only)/li> li>Lattice generation (both decoders)/li> li>Hypothesis files in NIST formats (SCLITE can be use for scoring hypotheses)/li> /ul> /li> li>Acoustic modeling ul> li>Acoustic models based on continuous density Hidden Markov Models (CD-HMMs) with emission probabilities modeled using mixtures of Gaussian distributions (GMMs)/li> li>HMM topology fixed to three states left to right/li> li>Variable number of Gaussian components per HMM-state/li> li>No explicit modeling of transition probabilities/li> li>Diagonal and full covariance modeling/li> li>Cross-word context dependency modeling using triphone, pentaphones, heptaphones, etc/li> li>Maximum Likelihood Estimation criterion/li> li>Discriminative training using boosted Maximum Mutual Information (bMMI) criterion with I-smoothing and cancellation of statistics/li> li>Parallel accumulation of sufficient statistics for both Maximum Likelihood and Discriminative Training criteria/li> li>Linear algebra support through template classes (Matrix, Vector, etc) wrapping third party libraries (BLAS and LAPACK)/li> /ul> /li> li>Language modeling ul> li>Support for n-gram language models in ARPA and binary formats/li> li>Support for any n-gram order (zerogram, unigram, bigram, trigram, fourgram, etc)/li> li>Language models are internally represented as Finite State Machines/li> /ul> /li> li>Speaker adaptation ul> li>Model space Maximum Likelihood Linear Regression (MLLR) using regression trees to automatically determine the number of transforms to be used and how adaptation data is shared among transforms/li> li>Feature space Maximum Likelihood Linear Regression (fMLLR)/li> li>Vocal Tract Length Normalization (VTLN)/li> /ul> /li> li>Feature extraction ul> li>Mel Frequency Cepstral Coefficients (MFCC) features/li> li>Cesptral Mean Normalization (CMN) and Cepstral Mean Variance Normalization (CMVN) at both utterance or session level/li> li>Feature decorrelation and dimensionality reduction using Heteroscedastic Linear Discriminant Analysis (HLDA)/li> li>Support for spliced features and third order derivatives/li> /ul> /li> li>Lattice processing and n-best list generation ul> li>Lattice rescoring using different criteria: maximum likelihood or posterior probabilities/li> li>N-best generation (from lattices) using different criteria: maximum likelihood or posterior probabilities/li> li>Lattice word error rate (WER) computation (oracle)/li> li>Lattice alignment and HMM-state marking/li> li>Attach LM-scores to lattice edges according to a given language model/li> li>Lattice-based posterior probability computation/li> li>Confidence annotation/li> li>Lattice path-insertion (discriminative training)/li> li>Lattices are processed in binary format but text format is available for readability purposes/li> /ul> /li> li> Speech activity detection ul> li>HMM-based speech activity detection/li> ul> /li>/ul>/div>/div>!-- end welcome -->div idlicense classstd>h1>License/h1>p>The Bavieca speech recognition toolkit is an open source project distributed under the highly unrestricted a href./files/LICENSE.txt>Apache 2.0 license/a>, and is freely available on SourceForge./p>/div>!-- end license -->div idbenchmarks classstd>h1>Benchmarks/h1>p>The recognition accuracy and real time performance of Bavieca has been measured on different tasks./p>h2>WSJ Nov92 Evaluation (microphone read speech)/h2>The table below summarizes the accuracy of Bavieca in the WSJ Nov92 task compared to known speech recognition systems. Additional details about this evaluation as well as data regarding the real-time performanceof Bavieca on this task can be found in the research article a href./articles/baviecaII.pdf>The Bavieca Speech Recognition Toolkit/a> presented at IEEE SLT 2012./p>table classtooltable>tr> th>/th>th colspan2>5k/th>th colspan2>20k/th>/tr>tr> th>system/th> th>bigram/th> th>trigram/th> th>bigram/th> th>trigram/th> th>gender dep./th> /tr>tr> td>HTK/td> td>5.1/td> td>3.2/td> td>11.1/td> td>9.5/td> td>yes/td> /tr>tr> td>Limsi/td> td>4.8/td> td>3.1/td> td>11.0/td> td>9.1/td> td>yes/td> /tr>tr> td>Kaldi/td> td>/td> td>/td> td>11.8/td> td>/td> td>no/td> /tr>tr> td>Bavieca /td> td>4.7/td> td>3.1/td> td>10.6/td> td>8.7/td> td>no/td> /tr>tr> td>Bavieca+bMMIsup>*/sup>/td> td>/td> td>2.8/td> td>/td> td>8.2/td> td>no/td> /tr>/table>b>Table./b> WER (%) on the WSJ Nov92 for Bavieca and other speech recognition toolkits.p>* This sysem configuration uses discriminatively trained acoustic models, so it is not directly comparable to the rest of the systems in the table, which use acoustic models trained under Maximum Likelihood./p>!-- h2>My Science Tutor (conversational children speech)/h2> -->/div>div idpublications classstd>h1>Publications/h1>p>Baviecas main publication is:/p>a href./articles/baviecaII.pdf>D. Bolaños, “The Bavieca Open-Source Speech Recognition Toolkit”. i>In Proceedings of IEEE Workshop on Spoken Language Technology (SLT)/i>, December 2-5, 2012, Miami, FL./a>p>Additionally, Bavieca has been utilized in several research projects, mainly in the education and speech recognition fields. Below there is a list of journal articles describing research that made use of the Bavieca speechrecogniton toolkit./p> ul> li> D. Bolaños, R. A. Cole, W. H. Ward, G. A. Tindal, J. Hasbrouck, P. J. Schwanenflugel. “Human and Automated Assessment of Oral Reading Fluency”, i>Journal of Educational Psychology, special issue on Advanced Learning Tecnologies/i>, 2012 (in press)./li>li> D. Bolaños, R. A. Cole, W. H. Ward, G. A. Tindal, P. J. Schwanenflugel, M. R. Kuhn. “Automatic Assessment of Expressive Oral Reading”, i>Speech Communication/i> (Volume 55, Issue 2, February 2013, Pages 221–236))./li>li> W. H. Ward, R. A. Cole, D. Bolaños, et al. “My Science Tutor, a Conversational Virtual Tutor for Teaching Science to Elementary Students”. i>Journal of Educational Psychology, special issue on Advanced Learning Tecnologies/i>, 2012 (accepted for publication, in press)./li>li> D. Bolaños, R. A. Cole, W. H. Ward, E. Borts and E. Svirsky. “FLORA, FLuent Oral Reading Assessment of Children’s Speech”, i>ACM Transactions on Speech and Language Processing, Special Issue on Speech and Language Processing of Children’s Speech for Child-machine Interaction Applications, August 2011./i>/li>li> W. H. Ward, R. A. Cole, D. Bolaños, C. Buchenroth-Martin, E. Svirsky, S. van Vuuren, T. Weston and J. Zheng. “My Science Tutor: A Conversational Multi-Media Virtual Tutor for Elementary School Science”, i>ACM Transactions on Speech and Language Processing, Special Issue on Speech and Language Processing of Children’s Speech for Child-machine Interaction Applications, August 2011./i>/li>/ul>/div>!-- end publications -->div iddownloadSub classstd>h1>a iddownloadLink>Download/a>/h1>p>Baviecas source code can be freely downloaded from SourceForge. The following command downloads Baviecas source code from the git repository at SourceForge:/p>p>code>git clone git://git.code.sf.net/p/bavieca/code bavieca-code/code>/p>/div>!-- end download -->div idinstall classstd>h1>a idinstallLink>Install/a>/h1>h2>Prerequisites/h2>p>Bavieca makes use of the BLAS and LAPACK third-party libraries to perform linear algebra, thus, it is necessaryto download and compile these libraries before installing Bavieca. These libraries can be obtained from a hrefhttp://www.netlib.org/>netlib/a>. Once the libraries are installed, some variables defined in code>/src/Makefile.defines/code> need to be modified to point to the installed libraries. The followingvariables need to be set:/p>ul> li>code>INCS_DIR_CBLAS /code> directory containing BLAS include files/li> li>code>INCS_DIR_LAPACK /code> directory containing LAPACK include files/li> li>code>LIBS_DIR_CBLAS /code> directory containing CBLAS library files/li> li>code>LIBS_DIR_LAPACK /code> directory containing LAPACK library files/li> li>code>LIB_CBLAS /code> CBLAS library files/li> li>code>LIB_LAPACK /code> LAPACK library files/li>/ul>h2>Installation/h2>p>Currently there are no binary distributions of Bavieca, thus it is necessary to compileBavieca for the target architecture to produce the actual binaries. Baviecas source code has been compiled under Linux (32 and 64 bits) and Windows (64 bits) x86 architectures. However,at the moment only some of its functionality (mainly the functionality exposed in Baviecas API) has been tested on Windows. For this reason the current version of the software in the repositorydoes not contain support for compiling it under Windows. Future releases will incorporate full support for the Windows platform and possibly for other platforms./p>p>Once Baviecas dependencies have been installed, it is possible to compile the toolkit by navigating to the directory code>/src/code> and typing code>make/code>. This commandwill produce an executable file for each command line tool in Bavieca and the Bavieca API library./p>/div>!-- end sourcecode -->div idprerequisites classstd>h1>a idprerequisitesLink>Prerequisites/a>/h1>/div>!-- end prerequisites -->script>$(document).ready(function(){ !-- show --> $(ul.downloadDrop).show(); $(#welcome).show(); !-- hide --> $(ul.downloadDrop).hide(); $(ul.installationDrop).hide(); $(ul.benchmarksDrop).hide(); $(#welcomeItem).click(function(){ $(div.std).hide(); $(#welcome).show(); }); $(#licenseItem).click(function(){ $(div.std).hide(); $(#license).show(); }); $(#benchmarksItem).click(function(){ $(div.std).hide(); $(#benchmarks).show(); }); $(#publicationsItem).click(function(){ $(div.std).hide(); $(#publications).show(); }); $(#downloadItem).click(function(){ $(div.std).hide(); $(#downloadSub).show(); }); $(#installItem).click(function(){ $(div.std).hide(); $(#install).show(); }); !-- links --> $(.publicationsShow).click(function() { $(div.std).hide(); $(#publications).show(); }); $(.benchmarksShow).click(function() { $(div.std).hide(); $(#benchmarks).show(); });});/script>/div>!-- end main -->!-- footer -->div idfooter>div idfooter_left>© Copyright 2012 Daniel Bolaños/div>div idfooter_right>/div>/div>!-- end footer -->/body>/html>
View on OTX
|
View on ThreatMiner
Please enable JavaScript to view the
comments powered by Disqus.
Data with thanks to
AlienVault OTX
,
VirusTotal
,
Malwr
and
others
. [
Sitemap
]