Adastra 2.0 - C# on top of OpenVibe

Come here to discuss about OpenViBE in general!
Post Reply
toncho11
Posts: 123
Joined: Tue Apr 19, 2011 7:58 pm

Adastra 2.0 - C# on top of OpenVibe

Post by toncho11 »

I have released Adastra 2.0. Adastra runs on top of OpenVibe. Adastra focuses mainly on machine learning over EEG signals. It is written in C# and it can be easily extended. Adastra can be used for BCI experiments such as brain controlled mouse cursor movement, etc.

http://code.google.com/p/adastra/

Please, try it and provide me with feedback. I am also searching for co-developers.

Thanks,
-Anton

toncho11
Posts: 123
Joined: Tue Apr 19, 2011 7:58 pm

Re: Adastra 2.0 - C# on top of OpenVibe

Post by toncho11 »

Adastra 2.2.1 has been released.

Check it out: http://code.google.com/p/adastra/downloads/list

toncho11
Posts: 123
Joined: Tue Apr 19, 2011 7:58 pm

Re: Adastra 2.0 - C# on top of OpenVibe

Post by toncho11 »

Version 2.5 has been released.

Adastra supports Logistic Regression through C# Octave wrapper. Octave scripts are part of the distribution. Other machine learning algorithms will also be implemented in Octave and added to Adastra. DSP filters in Matlab/Octave can also be added.

Enjoy :)

toncho11
Posts: 123
Joined: Tue Apr 19, 2011 7:58 pm

Re: Adastra 2.0 - C# on top of OpenVibe

Post by toncho11 »

Version 2.5.1 is a maintenance release.

It is important that one leave OpenVibe fully load before using Adastra. If not fully loaded Adastra might acquire corrupted data that will later prevent machine learning algorithms work correctly resulting in an exception. Several checks and warnings of this type have been added for this version. Also sufficient CPU power is required to both run OpenVibe and Adastra - no laptops on battery are recommended.

As always please uninstall your previous version and then make sure you are using version 2.5.1 from Menu -> Help -> About.

toncho11
Posts: 123
Joined: Tue Apr 19, 2011 7:58 pm

Re: Adastra 2.0 - C# on top of OpenVibe

Post by toncho11 »

Version 2.6 is available.

Adastra now considers the sampling frequency and produces signal chunks (or epochs). These signal chunks are used by the "EigenValuesFeatureGenerator" to calculate covariance matrix for each one. Then the eigen values of each covariance matrix are used as feature vectors.

This new feature generator is used in the Emotiv version, the OpenVibe scenario still uses the OpenVibe built in feature generator, but this will change in the future versions.

I will need to implement some UI for workflow configuration, probably in WPF.

-Anton

Post Reply