@inproceedings{MTMT:32813602,
	title = {An Adaptive Resonance Theory-based Neural Network for Autonomous Learning via Iterative Knowledge Redescription},
	url = {https://m2.mtmt.hu/api/publication/32813602},
	author = {Bártfai, Gusztáv},
	booktitle = {Proceedings of the International Joint Conference on Neural Networks},
	doi = {10.1109/IJCNN52387.2021.9533351},
	unique-id = {32813602},
	abstract = {This paper introduces a neural network architecture called ARTIR, which autonomously learns classifications of arbitrary input sequences via iterative redescription of knowledge that it builds up during learning. Knowledge redescription is achieved via changes to both the network architecture and the connection weights between neurons. Based on ARTMAP, the supervised version of the Adaptive Resonance Theory (ART) neural network, and motivated by the Representational Redescription hypothesis in cognitive science, the ARTIR network creates new, successively deeper internal features iteratively. The process is driven by the complexity of the classification task as measured by the number of input categories the ARTMAP network develops during training. The network is forced to discover a new internal feature that better discriminates the output classes every time the currently perceived difficulty of the learning task exceeds a certain limit as indicated by the number of input categories the ARTMAP network develops in response to an input-target sequence. Once a new candidate feature - found to correlate well enough with an output class - is accepted, ARTIR re-learns the mapping using the additional feature and uncommits its categories that are no longer needed as a result of re-learning. The iterative learning process then continues where the newly created feature, along with all original inputs and internal features created during previous stages of learning, becomes subject of internal knowledge redescription for future iterations. Experiments carried out on two binary decision problems demonstrate the viability of the approach, and the results point to further questions that will be the subject of future research.},
	keywords = {Neural network; art; Computer Science, Hardware & Architecture; Computer Science, Artificial Intelligence; Representation learning; adaptive resonance theory; ARTMAP; ARTMAP; Representational redescription; ARTIR; knowledge redescription; incremental feature discovery; autonomous adaptive systems; continual learning},
	year = {2021}
}

@article{MTMT:3003486,
	title = {FPGAs bringen Vorteile gegenüber DSPs},
	url = {https://m2.mtmt.hu/api/publication/3003486},
	author = {Rekeczky, Csaba and Bártfai, Gusztáv and Bob, Siller},
	journal-iso = {ELECTRON IND},
	journal = {ELEKTRONIK INDUSTRIE},
	volume = {2012},
	unique-id = {3003486},
	issn = {0174-5522},
	year = {2012},
	pages = {18-21}
}

@article{MTMT:165614,
	title = {Smart sensor based vision system for automated processes.},
	url = {https://m2.mtmt.hu/api/publication/165614},
	author = {Argyros, A and Bártfai, Gusztáv and Eitzinger, C and Kemény, Zsolt and Csáji, Balázs Csanád and Kék, László and Lourakis, M and Reisner, W and Sandrisser, W and Sarmis, T and Umgeher, G and Viharos, Zsolt János},
	journal-iso = {INT J FACT AUTOM ROBOT SOFT COMPUT},
	journal = {INTERNATIONAL JOURNAL OF FACTORY AUTOMATION ROBOTICS AND SOFT COMPUTING},
	volume = {3},
	unique-id = {165614},
	issn = {1828-6984},
	year = {2007},
	pages = {118-123},
	orcid-numbers = {Kemény, Zsolt/0000-0003-3866-6556; Viharos, Zsolt János/0000-0002-9561-6857}
}

@inbook{MTMT:165481,
	title = {Smart sensor based vision system for automated processes.},
	url = {https://m2.mtmt.hu/api/publication/165481},
	author = {Argyros, A and Bártfai, Gusztáv and Eitzinger, C and Kemény, Zsolt and Csáji, Balázs Csanád and Kék, László and Lourakis, M and Reisner, W and Sandrisser, W and Sarmis, T and Umgeher, G and Viharos, Zsolt János},
	booktitle = {Emerging technologies, robotics and control systems. Vol. 2},
	unique-id = {165481},
	year = {2007},
	pages = {24-29},
	orcid-numbers = {Kemény, Zsolt/0000-0003-3866-6556; Viharos, Zsolt János/0000-0002-9561-6857}
}

@article{MTMT:1185448,
	title = {Fault-tolerant design of analogic CNN templates and algorithms - Part I: The binary output case},
	url = {https://m2.mtmt.hu/api/publication/1185448},
	author = {Földesy, Péter and Kék, László and Zarándy, Ákos and Roska, Tamás and Bártfai, Gusztáv},
	doi = {10.1109/81.747209},
	journal-iso = {IEEE T CIRC SYST FUND},
	journal = {IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I - FUNDAMENTAL THEORY AND APPLICATIONS},
	volume = {46},
	unique-id = {1185448},
	issn = {1057-7122},
	abstract = {This paper addresses the issue of designing a class of fault-tolerant
		cellular neural network (CNN) templates that, combined with CNN
		analogic algorithms, work correctly and reliably on given CNN universal
		machine (CNN-UM) chips. In particular, a generic method for finding
		nonpropagating binary-output CNN templates is proposed, This method is
		based on measurements of actual CNN-UM chips and combines adaptive
		optimization and decomposition of theoretically ideal CNN templates in
		order to correct the erroneous behavior of actual CNN-UM chips, which
		is mainly caused by imperfections introduced during fabrication, More
		specifically, the entire array of cells in a CNN-UM chip is modeled by
		a single feed-forward virtual cell whose optimal parameters are found
		by a simple and effective gradient-based method. In the case of binary
		input-output uncoupled templates (or Boolean operators), a systematic
		template decomposition method is applied whenever optimization fails to
		find a correctly working CNN template for all possible combinations of
		local 3 x 3 binary input patterns. The resulting templates are finally
		combined, yielding a simple CNN analogic algorithm. Examples are
		presented for both binary- and analog-input operators, using two
		concrete stored-program CNN-UM chips to demonstrate the effectiveness
		of the proposed method, whose advantages and limitations are also
		discussed.},
	year = {1999},
	pages = {312-322},
	orcid-numbers = {Földesy, Péter/0000-0001-7495-0971}
}

@mastersthesis{MTMT:161879,
	title = {Towards implementation of biologically inspired autonomous and adaptive information processing systems},
	url = {https://m2.mtmt.hu/api/publication/161879},
	author = {Bártfai, Gusztáv},
	unique-id = {161879},
	year = {1999}
}

@techreport{MTMT:161870,
	title = {Combination of adaptive resonance theory (ART) and cellular neural network (CNN) chips into a high-speed pattern recognition system.},
	url = {https://m2.mtmt.hu/api/publication/161870},
	author = {Bártfai, Gusztáv},
	unique-id = {161870},
	year = {1999},
	pages = {1-16}
}

@inproceedings{MTMT:161590,
	title = {Fault tolerant CNN template design and optimatization based on chip measurements},
	url = {https://m2.mtmt.hu/api/publication/161590},
	author = {Földesy, Péter and Kék, László and Roska, Tamás and Zarándy, Ákos and Bártfai, Gusztáv},
	booktitle = {1998 fifth IEEE International Workshop on Cellular Neural Networks and Their Applications},
	doi = {10.1109/CNNA.1998.685415},
	unique-id = {161590},
	year = {1998},
	pages = {404-409},
	orcid-numbers = {Földesy, Péter/0000-0001-7495-0971}
}

@techreport{MTMT:161429,
	title = {Fault tolerant design of analogic CNN templates and algorithms. Part I: The binary output case.},
	url = {https://m2.mtmt.hu/api/publication/161429},
	author = {Földesy, Péter and Kék, László and Zarándy, Ákos and Roska, Tamás and Bártfai, Gusztáv},
	unique-id = {161429},
	year = {1998},
	pages = {1-18},
	orcid-numbers = {Földesy, Péter/0000-0001-7495-0971}
}

@article{MTMT:1001439,
	title = {A digital multiprocessor hardware accelerator board for Cellular Neural Networks: CNNHAC},
	url = {https://m2.mtmt.hu/api/publication/1001439},
	author = {Roska, Tamás and Bártfai, Gusztáv and Szolgay, Péter and Szirányi, Tamás and Radványi, András and Kozek, T and Ugray, Zs and Zarándy, Ákos},
	doi = {10.1002/cta.4490200512},
	journal-iso = {INT J CIRC THEOR APP},
	journal = {INTERNATIONAL JOURNAL OF CIRCUIT THEORY AND APPLICATIONS},
	volume = {20},
	unique-id = {1001439},
	issn = {0098-9886},
	year = {1992},
	eissn = {1097-007X},
	pages = {589-601},
	orcid-numbers = {Szirányi, Tamás/0000-0003-2989-0214}
}