Publications

sort by:  author date

Zpracování a kategorizace česky psaných textových dokumentů neuronovou sítí

MAUTNER, P., MOUČEK, R. Zpracování a kategorizace česky psaných textových dokumentů neuronovou sítí. In Informatika v škole a v praxi. Ružomberok: Pedagogická fakulta, 2008. s. 51-59. ISBN: 978-80-8084-362-5
Abstract PDF BibTeX

This article deals with application of WEBSOM methods for czech-written documnts collections. The basic principles of WEBSOM methods, transformation of text information into the real components feature vector and results of documents classification are described in the article. The Carpenter-Grossberg ART-2 neural network, normally used for adaptive vector clustering, was also tested as a document categorization tool. The results achieved by using of this method are also presented here.

EEG/ERP experiments - data and metadata structure

MOUČEK, R., JEŽEK, P. EEG/ERP experiments - data and metadata structure . In Neuroinformatics 2008. Stockholm : Frontiers , 2008. s. 123-123.
Abstract PDF BibTeX

The results of the EEG/ERP experiments (data, metadata)have to be stored to enable further analysis and interpretation. The first version of general EEG/ERP data format nad metadata structures was proposed.

Attention of drivers and dual task performance

MOUČEK, R., PODLENA, V. Attention of drivers and dual task performance. In Xth International conference on cognitive neuroscience. Istanbul : Istanbul University , 2008. s. 365-365.
Abstract PDF BibTeX

Performed ERP experiment is a variation on attentional blink paradigm; audio neose is added in control experiment. Software application (XML, Java) was developed. There is significant decrease in N400 amplitude in control experiment.

Combining two different signals for BCI

HOŠNA, M., MOUČEK, R. Combining two different signals for BCI. In Xth International conference on Cognitive Neuroscience. Bodrum: Lundbeck Ilac Ticaret Limited Sirketi, 2008. s. 1-1.
Abstract PDF BibTeX

One problem during the development of the brain computer interface is the fatigue. The fatigue may cause a performance drop and also raise the number of false choices. In this article we present combination of two different signal commonly used in BCI: the P3 component of the evoked potential and the steady-state visual evoked potential. Purpose of this combination is to improve the performance and also to extend the time how long a subject is able to control a BCI.

Detection of ERP using Matching Pursuit algorithm

SVOBODA, J., MAUTNER, P., MOUČEK, R. Detection of ERP using Matching Pursuit algorithm. In Xth international conference on cognitive neuroscience. Bodrum: Lundbeck Ilac Ticaret Limited Sirketi, 2008. s. 226.
Abstract PDF BibTeX

This article deals with application of matching pursuit algorithm for ERP processing. In matching pursuit method, the given signal can be approximated or represented as a superposition of waveforms that belong to redundant dictionary of functions. The waveforms are selected in order to best match the signal structure. This method can be successfully used for detection of the averaged ERP waves. Alternatively, the algorithm can be also used for detection of evoked response in non-averaged signal with aim to minimize ERP detection time.

Our partners