Modular Audio Recognition Framework News

I got a missing files report in Alpha 2 release of MARFCAT 0.0.1 from SATE 2010; this has been corrected in the Alpha 3 version.

(A new updated version is being developed for future release with SATE IV).

-s

25023 as of this post and growing! :-)

http://sourceforge.net/project/stats/detail.php?group_id=63118&ugn=marf&type=prdownload&mode=alltime&package_id=0

-s

MARF as a PureData extension is presented here:

http://artengine.ca/~catalogue-pd/32-Mokhov.pdf

the code of the PD externals will be put in the
CVS and released in FRS soon.

More about PureData:

http://puredata.org/

-s

As of this writing it is 20,177 downloads! :-)

Download stats are here:

https://sourceforge.net/project/stats/detail.php?group_id=63118&ugn=marf&type=prdownload&mode=alltime&package_id=0

-s

The subject says it all :-) As of this
post it is 15,226. And the month of March
was highest number of download ever: 1,339,
in one month!

-s

On March 14, The MARF Research and Development
group got its official PEN SMI number from IANA,
which is 28218.

http://www.iana.org/assignments/enterprise-numbers

This is done as a part of experimental ongoing
project of SNMPv2/SMIv2 network management of
the Distributed MARF nodes. This is a subbranch
unde the Distributed MARF's branch, currently
in its beginnings of development.

-s

In mid-March the CVS of Distributed MARF branch
was integrated into the main SF CVS repo of MARF
with the rest of stuff. It was made compile
with Eclipse and things got re-arranged a bit
from the PoC-demo release, so it can now
benefit from the mainstream MARF fixes and
further development under revision control.

For the developers and anonymous checkouts
from the CVS to get the distributed brach's
source code, checkout this branch:... read more

The changes in this release are as follows:
A bug has been fixed in Sample.setSampleSize() where the if()
condition would always evaluate to false when checking whether the
resulting sample is shorter or longer than the original. A bug in
custom Sample serialization where the order of writing an int and the
array data was reversed has been fixed. The comments have been updated
a bit more for the updates. Some related unit tests have been
added/fixed in order for the failures not to reoccur. The marf
mini-app's purpose has been clarified.... read more

Just noticed MARF surpassed 10000 mark in the
downloads section that were done through SF's
File Release System (FRS). That does not count
direct downloads from the site or mirrors.
This is quite an achievment for the project
at this time!

:-)

This CVS snaphot release is primarily a bug-fix
release. See the details below.

Changes:

* Fix the bug #1556320 of FeatureExtractionAggregator
dealing with NullPointerException's generated by
the aggreated modules. Since the introduction of
the new API of extractFeatures(double[]),
handling of its parameter was not done properly
when either querying a Preprocessing module or
acting directly on the parameter was reversed,
and, also a null in one case. Make it correct
and for the parameter wrap it into Raw, so the
handling is always consistent with a non-null
Preprocessing. In the process of debugging this,
enhance error reporting.... read more

This a new CVS snaphost release feature a number of
bug fixes and completion of integration of a couple
feature requests. While being the latest and the
greatest bleeding edge code release, it may also
contain latest bugs. It passes well though the
standard testing runs. The details on the changes
are provided below in the ChangeLog. Any constructive
feedback is welcome.

Download:
https://sourceforge.net/project/showfiles.php?group_id=63118&release_id=446448... read more

This release of the Zipf's Law testing application
is primarily a bugfix release fix a couple issues.
See the ChangeLog [1] for details. Download at [2].

[1] http://sourceforge.net/project/shownotes.php?release_id=444869&group_id=63118
[2] http://sourceforge.net/project/showfiles.php?group_id=63118&package_id=153614&release_id=444869

A new CVS snapshot is out. It's primarily featuring the following improvements:

* Preprocessing: implement silence removal
* Preprocessing: fix bug #1550183 in Endpoint to do
normalization properly

Notice, snaphot repleases may contain latest-greatest
features as well as bugs.

You can download it from SourceForge.net
by following this link:

https://sourceforge.net/project/showfiles.php?group_id=63118&release_id=444313

Announcement of the initial release of the Distributed MARF
proof-of-concept (PoC) demo implementation.

MARF's pipeline stages can now run on different
computers in more than one number using three
different communication technologies in the PoC.

The technologies include:

* CORBA
* Java RMI
* Web Services

The six implemented services include:

* SpeakerIdent Front-end Service (plus the client),
which are mods from the classical SpeakerIdentApp... read more

New snapshot with the following changes:

* Update Version to report the build as well in the
PACKAGE constant.

* Add a new API to ISampleLoaders and its implementation
to load from byte array loadSample(byte[]) and a stream
loadSample(InputStream) as convenience methods and
convert most implemented ones to use the streamed
version. The byte array is needed for Distributed MARF
as the preliminary implementation will be passing the
file over as a byte array over the network.... read more

The new snapshot contains the following major changes:

* Classification: IClassification and Classification got
new API of train(double[]) and classify(double[]) for
flexibility and ability to use the modules remotely
quicker.

* Feature extraction: IFeatureExtraction and
FeatureExtraction got a new API of
extractFeatures(double[]) pretty much for the same
reasons as for Classification.... read more

This next development snapshot among many changes includes:

* Fixes of bugs 1515258 and 1508630.

* Enables NeuralNetwork to dump its data not only in XML,
but also in DUMP_GZIP_BINARY and DUMP_BINARY formats.

* Refactoring of Classification and TrainingSet and
RandomClassification to take the advantage of of the
general implementation of StorageManager's dump() and
restore() API.

* General peformance improvements related to the use
of StringBuffer insted of String concatenation.... read more

This is a current development snapshot release
of the fresh CVS checkout of the 'marf' module
and its compiled 'fat' jar. It contains all
latest features and fixes ... and perhaps bugs.
For the folks who want to try out stuff and
don't use CVS for that.

http://sourceforge.net/project/showfiles.php?group_id=63118&package_id=197804&release_id=433598

-s

TestFFT [1], TestLPC [2], TestLoaders [3], and TestWaveloader [4]
apps of project MARF [5] were released. This is a minor
maintenance release for the apps to be consistent
with all the previously released. Primarily
directed at folks not using the CVS to get the
updates. All apps work with MARF 0.3.0.5 properly.

TestFilters [6] was not released as a part of this
maintenance bundle because of some ongoing
work requiring MARF 0.3.0.6. A maintenance
release of TestFilters can be made upon request.
Else will wait till the 0.3.0-final.... read more

After SF.net have corrected and redeployed their
CVS service after a lengthy outage, MARF{s pages
and online API docs are updated regularly again.

The 0.3.0-devel-20060226 (0.3.0.5) release is a maintenance release with several bug fixes and new features of MARF. MARF is a general modular cross-platform framework with a collection of algorithms for voice and natural language text analysis and recognition along with sample applications of its use, implemented in Java.

A new set of applications are released with the new MARF, including applications that employ/test some NLP techniques -- ZipfLaw, LangIdentApp, and ProbabilisticParsingApp. Additionally, TestPlugin application to test plug-in infrastructure of MARF and TestNN for NeuralNetwork test are first publicly released. MathTestApp and SpeakerIdentApp also experienced a release due to new features and bug fixes. Endpointing got implemented for Preprocessing and more code refactoring went in.... read more

Just a quick update that .md5 checksums for each
previously released files (except for 0.1.1 and
0.1.2 releases) are added if anyone wants to
verify them after download.

All future releases will have it at the time of
the release.

-s

Hello all,

I posted static version of older versions of MARF on the web page, so one can easier consult them, especially looking for API differences between releases w/o the necessity of downloading all the doc tarballs.

http://marf.sourceforge.net/#alldocs

-s

The 0.3.0-devel-20050817 release is a mostly maintenance release with a few bug fixes and new features of MARF. MARF is a general modular cross-platform framework with a collection of algorithms for audio (voice, speech, and sound) and natural language text analysis and recognition along with sample applications of its use, implemented in Java.

The bug fixes touch the utility modules marf.util.OptionProcessor and marf.util.Arrays (see the website http://marf.sf.net for details). The new features include addition of the plug-in interface for filters IFilter and overall completion of the plug-in framework for the preprocessing, feature extraction, and classification module types. OptionProcessors got a new API of size() and getArgumentVector(). FFT came more customizable through the class interface and ModuleParams facility. All concrete Serializable classes got an explicit serialVersionUID, as recommended by the Java specification. For the rest of the changes, additions, and improvements please see the Release Notes and ChangeLog. ... read more

The 0.3.0-devel-20050730 release is a mostly maintenance release with a few bug fixes, but also provides a few new features and applications of MARF. MARF is a general modular cross-platform framework with a collection of algorithms for audio (voice, speech, and sound) and natural language text analysis and recognition along with sample applications (identification, NLP, etc.) of its use, implemented in Java.... read more