We describe nonnegative matrix factorisation (NMF) with a Kullback-Leibler (KL) error measure in a statistical
framework, with a hierarchical generative model consisting of an observation and a prior component. Omitting the prior
leads to the standard KL-NMF algorithms as special cases, where maximum likelihood parameter estimation is carried
out via the Expectation-Maximisation (EM) algorithm. Starting from this view, we develop full Bayesian inference
via variational Bayes or Monte Carlo. Our construction retains conjugacy and enables us to develop more powerful
models while retaining attractive features of standard NMF such as monotonic convergence and easy implementation.
We illustrate our approach on model order selection and image reconstruction.

Ali Taylan Cemgil

Bayesian Inference for Nonnegative Matrix Factorisation Models

We discuss the BCI based on inner tones and inner music. We had some success in the detection of inner tones, the imagined tones which are not sung aloud. Rather easily imagined and controlled, they offer a set of states usable for BCI, with high information capacity and high transfer rates. Imagination of sounds or musical tunes could provide a multicommand language for BCI, as if using the natural language. Moreover, this approach could be used to test musical abilities. Such BCI interface could be superior when there is a need for a broader command language. Some computational
estimates and unresolved difficulties are presented.

Wlodzimierz Klonowski

Wlodzis&#x142;aw Duch

Aleksandar Perovic

Aleksandar Jovanovic

Some Computational Aspects of the Brain Computer Interfaces Based on Inner Music

This paper demonstrates that nonnegative matrix factorisation is mathematically related to a class
of neural networks that employ negative feedback as a mechanism of competition. This observation
inspires a novel learning algorithm which we call Divisive Input Modulation (DIM). The proposed
algorithm provides a mathematically simple and computationally efficient method for the unsupervised
learning of image components, even in conditions where these elementary features overlap
considerably. To test the proposed algorithm, a novel artificial task is introduced which is similar
to the frequently-used bars problem but employs squares rather than bars to increase the degree of
overlap between components. Using this task, we investigate how the proposed method performs on
the parsing of artificial images composed of overlapping features, given the correct representation
of the individual components; and secondly, we investigate how well it can learn the elementary
components from artificial training images. We compare the performance of the proposed algorithm
with its predecessors including variations on these algorithms that have produced state-of-the-art
performance on the bars problem. The proposed algorithm is more successful than its predecessors
in dealing with overlap and occlusion in the artificial task that has been used to assess performance.

M. W. Spratling

K. De Meyer

R. Kompass

Unsupervised Learning of Overlapping Image Components Using Divisive Input Modulation

EEG-based discrimination between different motor imagery states has been subject of a number of studies in healthy subjects. We investigated the EEG of 15 patients with complete spinal cord injury during imagined right hand, left hand, and feet movements. In detail we studied pair-wise discrimination functions between the 3 types of motor imagery. The following classification accuracies (mean <svg     style="vertical-align:-0.3003pt;width:10.1px;" id="M1" height="9.5625" version="1.1" viewBox="0 0 10.1 9.5625" width="10.1"  xmlns:xlink="http://www.w3.org/1999/xlink" xmlns="http://www.w3.org/2000/svg">
	
		
			<g transform="matrix(.017,-0,0,-.017,.062,9.125)"><path id="xB1" d="M535 290h-212v-206h-58v206h-213v50h213v193h58v-193h212v-50zM535 -22h-483v50h483v-50z" /></g>

		
	
</svg> SD) were obtained: left versus right hand <svg     style="vertical-align:-0.4368pt;width:101.0375px;" id="M2" height="11.75" version="1.1" viewBox="0 0 101.0375 11.75" width="101.0375"  xmlns:xlink="http://www.w3.org/1999/xlink" xmlns="http://www.w3.org/2000/svg">
	
		
			<g transform="matrix(.017,-0,0,-.017,.062,11.15)"><path id="x36" d="M137 343l67 33q37 17 63 17q79 0 129.5 -53t50.5 -131q0 -92 -58 -156.5t-147 -64.5t-147 68t-58 182q0 63 17 119t43 95.5t61.5 72t69 52t67.5 31.5q62 22 128 33l6 -32q-56 -11 -108 -35q-149 -71 -184 -231zM227 337q-47 0 -95 -27q-6 -23 -6 -70q0 -93 36 -155.5&#xA;t96 -62.5q53 0 78 45.5t25 105.5q0 68 -35 116t-99 48z" /></g><g transform="matrix(.017,-0,0,-.017,8.222,11.15)"><path id="x35" d="M153 550l-26 -186q79 31 111 31q90 0 141.5 -51t51.5 -119q0 -93 -89 -166q-85 -69 -173 -71q-32 0 -61.5 11.5t-41.5 23.5q-18 17 -17 34q2 16 22 33q14 9 26 -1q61 -50 124 -50q60 0 93 43.5t33 104.5q0 69 -41.5 110t-121.5 41q-53 0 -102 -20l38 305h286l6 -8&#xA;l-26 -65h-233z" /></g><g transform="matrix(.017,-0,0,-.017,16.381,11.15)"><path id="x2E" d="M113 -12q-24 0 -39.5 16t-15.5 42q0 24 16 40.5t40 16.5t40 -16.5t16 -40.5q0 -26 -16 -42t-41 -16z" /></g><g transform="matrix(.017,-0,0,-.017,20.257,11.15)"><path id="x30" d="M241 635q53 0 94 -28.5t63.5 -76t33.5 -102.5t11 -116q0 -58 -11 -112.5t-34 -103.5t-63.5 -78.5t-94.5 -29.5t-95 28t-64.5 75t-34.5 102.5t-11 118.5q0 58 11.5 112.5t34.5 103t64.5 78t95.5 29.5zM238 602q-32 0 -55.5 -25t-35.5 -68t-17.5 -91t-5.5 -105&#xA;q0 -76 10 -138.5t37 -107.5t69 -45q32 0 55.5 25t35.5 68.5t17.5 91.5t5.5 105t-5.5 105.5t-18 92t-36 68t-56.5 24.5z" /></g><g transform="matrix(.017,-0,0,-.017,28.416,11.15)"><path id="x33" d="M285 378v-2q65 -13 102 -54.5t37 -97.5q0 -57 -30.5 -104.5t-74 -75t-85.5 -42t-72 -14.5q-31 0 -59.5 11t-40.5 23q-19 18 -16 36q1 16 23 33q13 10 24 0q58 -51 124 -51q55 0 88 40t33 112q0 64 -39 96.5t-88 32.5q-29 0 -64 -11l-6 29q77 25 118 57.5t41 84.5&#xA;q0 45 -26.5 69.5t-68.5 24.5q-67 0 -120 -79l-20 20l43 63q51 56 127 56h1q66 0 107 -37t41 -95q0 -42 -31 -71q-22 -23 -68 -54z" /></g><g transform="matrix(.017,-0,0,-.017,36.576,11.15)"><path id="x25" d="M594 629l-404 -661l-31 23l402 661zM194 635q70 0 107 -56t37 -125t-37.5 -125.5t-107.5 -56.5t-107 56.5t-37 124.5q0 70 37.5 126t107.5 56zM193 605q-28 0 -45 -29t-21.5 -59.5t-4.5 -62.5t4.5 -63t22 -60t45.5 -29q22 0 37.5 16t22 42t9 48t2.5 45q0 152 -72 152z&#xA;M567 351q69 0 106 -56t37 -125t-37 -125.5t-107 -56.5t-107.5 56.5t-37.5 124.5q0 43 14.5 83t49 69.5t82.5 29.5zM566 322q-28 0 -45 -28.5t-22 -59.5t-5 -64q0 -23 3 -46t10.5 -48.5t23 -41.5t36.5 -16q28 0 45 29.5t21 59.5t4 62q0 24 -2.5 46.5t-9.5 48.5t-22 42t-37 16&#xA;z" /></g><g transform="matrix(.017,-0,0,-.017,56.022,11.15)"><use xlink:href="#xB1"/></g><g transform="matrix(.017,-0,0,-.017,72.613,11.15)"><path id="x38" d="M249 635q70 0 116 -43t46 -105q0 -46 -28 -80q-22 -25 -80 -64q62 -35 97 -75t35 -99q0 -81 -63 -131t-135 -50q-83 0 -137.5 45.5t-54.5 123.5q0 52 45 95q29 28 89 64q-109 62 -109 155q0 66 50.5 115t128.5 49zM238 603q-42 0 -67.5 -31t-25.5 -72q0 -50 32.5 -79.5&#xA;t98.5 -62.5q61 48 61 124q0 59 -29.5 90t-69.5 31zM248 20q46 0 76.5 33.5t30.5 89.5q0 50 -39 85.5t-110 71.5q-81 -54 -81 -137q0 -67 35.5 -105t87.5 -38z" /></g><g transform="matrix(.017,-0,0,-.017,80.773,11.15)"><use xlink:href="#x2E"/></g><g transform="matrix(.017,-0,0,-.017,84.648,11.15)"><use xlink:href="#x35"/></g><g transform="matrix(.017,-0,0,-.017,92.808,11.15)"><path id="x32" d="M412 140l28 -9q0 -2 -35 -131h-373v23q112 112 161 170q59 70 92 127t33 115q0 63 -31 98t-86 35q-75 0 -137 -93l-22 20l57 81q55 59 135 59q69 0 118.5 -46.5t49.5 -122.5q0 -62 -29.5 -114t-102.5 -130l-141 -149h186q42 0 58.5 10.5t38.5 56.5z" /></g>

		
	
</svg>, left hand versus feet <svg     style="vertical-align:-0.4368pt;width:109.2px;" id="M3" height="11.75" version="1.1" viewBox="0 0 109.2 11.75" width="109.2"  xmlns:xlink="http://www.w3.org/1999/xlink" xmlns="http://www.w3.org/2000/svg">
	
		
			<g transform="matrix(.017,-0,0,-.017,.062,11.15)"><use xlink:href="#x36"/></g><g transform="matrix(.017,-0,0,-.017,8.222,11.15)"><use xlink:href="#x38"/></g><g transform="matrix(.017,-0,0,-.017,16.381,11.15)"><use xlink:href="#x2E"/></g><g transform="matrix(.017,-0,0,-.017,20.257,11.15)"><path id="x31" d="M384 0h-275v27q67 5 81.5 18.5t14.5 68.5v385q0 38 -7.5 47.5t-40.5 10.5l-48 2v24q85 15 178 52v-521q0 -55 14.5 -68.5t82.5 -18.5v-27z" /></g><g transform="matrix(.017,-0,0,-.017,28.416,11.15)"><path id="x39" d="M244 635q90 0 143 -72t53 -177q0 -133 -65 -229.5t-171 -139.5q-79 -32 -140 -32l-5 30q109 18 185 91t101 186l-68 -36q-29 -16 -60 -16q-79 0 -129 51.5t-50 130.5q0 80 57 146.5t149 66.5zM228 602q-52 0 -78 -45.5t-26 -98.5q0 -69 36.5 -115.5t97.5 -46.5&#xA;q53 0 90 28q4 31 4 66q0 51 -9.5 95.5t-39 80.5t-75.5 36z" /></g><g transform="matrix(.017,-0,0,-.017,36.576,11.15)"><use xlink:href="#x25"/></g><g transform="matrix(.017,-0,0,-.017,56.022,11.15)"><use xlink:href="#xB1"/></g><g transform="matrix(.017,-0,0,-.017,72.613,11.15)"><use xlink:href="#x31"/></g><g transform="matrix(.017,-0,0,-.017,80.773,11.15)"><use xlink:href="#x31"/></g><g transform="matrix(.017,-0,0,-.017,88.932,11.15)"><use xlink:href="#x2E"/></g><g transform="matrix(.017,-0,0,-.017,92.808,11.15)"><use xlink:href="#x30"/></g><g transform="matrix(.017,-0,0,-.017,100.967,11.15)"><use xlink:href="#x38"/></g>

		
	
</svg>, and right hand versus feet <svg     style="vertical-align:-0.4368pt;width:101.0375px;" id="M4" height="11.75" version="1.1" viewBox="0 0 101.0375 11.75" width="101.0375"  xmlns:xlink="http://www.w3.org/1999/xlink" xmlns="http://www.w3.org/2000/svg">
	
		
			<g transform="matrix(.017,-0,0,-.017,.062,11.15)"><use xlink:href="#x36"/></g><g transform="matrix(.017,-0,0,-.017,8.222,11.15)"><use xlink:href="#x35"/></g><g transform="matrix(.017,-0,0,-.017,16.381,11.15)"><use xlink:href="#x2E"/></g><g transform="matrix(.017,-0,0,-.017,20.257,11.15)"><use xlink:href="#x30"/></g><g transform="matrix(.017,-0,0,-.017,28.416,11.15)"><use xlink:href="#x35"/></g><g transform="matrix(.017,-0,0,-.017,36.576,11.15)"><use xlink:href="#x25"/></g><g transform="matrix(.017,-0,0,-.017,56.022,11.15)"><use xlink:href="#xB1"/></g><g transform="matrix(.017,-0,0,-.017,72.613,11.15)"><use xlink:href="#x39"/></g><g transform="matrix(.017,-0,0,-.017,80.773,11.15)"><use xlink:href="#x2E"/></g><g transform="matrix(.017,-0,0,-.017,84.648,11.15)"><use xlink:href="#x32"/></g><g transform="matrix(.017,-0,0,-.017,92.808,11.15)"><use xlink:href="#x35"/></g>

		
	
</svg>. In 5 out of 8 paralegic patients, the discrimination accuracy was greater than 70&#x25; but in only 1 out of 7 tetraplagic patients. The present findings provide evidence that in the majority of paraplegic patients an EEG-based BCI could achieve satisfied results. In tetraplegic patients, however, it is expected that extensive training-sessions are necessary to achieve a good BCI performance at least in some subjects.

G. Pfurtscheller

P. Linortner

R. Winkler

G. Korisek

G. Müller-Putz

Discrimination of Motor Imagery-Induced EEG Patterns in Patients with Complete Spinal Cord Injury

In this paper, we present, with particular focus on the adopted processing and identification chain and protocol-related solutions, a whole self-paced brain-computer interface system based on a 4-class steady-state visual evoked potentials (SSVEPs) paradigm. The proposed system incorporates an automated spatial filtering technique centred on the common spatial patterns (CSPs) method, an autoscaled and effective signal features extraction which is used for providing an unsupervised biofeedback, and a robust self-paced classifier based on the discriminant analysis theory. The adopted operating protocol is structured in a screening, training, and testing phase aimed at collecting user-specific information regarding best stimulation frequencies, optimal sources identification, and overall system processing chain calibration in only a few minutes. The system, validated on 11 healthy/pathologic subjects, has proven to be reliable in terms of achievable communication speed (up to 70 bit/min) and very robust to false positive identifications.

Sergio Parini

Luca Maggi

Anna C. Turconi

Giuseppe Andreoni

A Robust and Self-Paced BCI System Based on a Four Class SSVEP Paradigm: Algorithms and Protocols for a High-Transfer-Rate Direct Brain Communication

About this journal

<p><em>Computational Intelligence and Neuroscience</em> is a forum for the interdisciplinary field of neural computing, neural engineering and artificial intelligence. The journal’s focus is on intelligent systems for computational neuroscience.</p>


Journal profile

Meet the editorial board

<p>Chief Editor, Professor Cichocki, engages in world-leading research in the field of artificial intelligence and biomedical applications of advanced data analytics technologies.</p>


Editor spotlight

Submit your proposal

Do you think there is an emerging area of research that really needs to be highlighted? Or an existing research area that has been overlooked or would benefit from deeper investigation? Raise the profile of a research area by leading a Special Issue.

Special Issues

Author guidelines

Editorial board

Databases and indexing

Computational Intelligence and Neuroscience - Table of contents 2009

Computational Intelligence and Neuroscience | Hindawi

title

Computational Intelligence and Neuroscience is a forum for the interdisciplinary field of neural computing, neural engineering and artificial intelligence. The journal’s focus is on intelligent systems for computational neuroscience.

description

Computational Intelligence and Neuroscience

https://images.hindawi.com/journals/cin/cin.banner.jpg

Computational Intelligence and Neuroscience