论文信息 - Learning connectedness in binary images

Learning connectedness in binary images

This paperproposesa new Eye-basedRecurrentNetwork Architecture(ERNA) for imageclassification. The new architectureis trainedby a combinationof QlearningandRPROP. Theclassificationperformanceis comparedwith othernetwork architecturesonthetaskof determiningconnectedness betweenpixelsin smallbinary images.Theexperimentsshow thatERNA outperformsboththestandardmulti-layer perceptronnetwork andthefully-connectedrecurrentnetwork on thetaskmentioned above. Thisperformanceleadsusto theconclusionthattheeye facilitateslearningin thetopologically-structuredomainof imageclassification.

[1] Marvin Minsky,et al. Perceptrons: An Introduction to Computational Geometry , 1969 .

[2] Martin A. Riedmiller,et al. A direct adaptive method for faster backpropagation learning: the RPROP algorithm , 1993, IEEE International Conference on Neural Networks.

[3] Jing Peng,et al. Incremental multi-step Q-learning , 1994, Machine Learning.

[4] J. Mycielski. Review: Marvin Minsky and Seymour Papert, Perceptrons, An Introduction to Computational Geometry , 1972 .

[5] Peter Dayan,et al. Technical Note: Q-Learning , 2004, Machine Learning.

[6] DeLiang Wang,et al. On Connectedness: A Solution Based on Oscillatory Correlation , 2000, Neural Computation.

[7] Barak A. Pearlmutter. Gradient calculations for dynamic recurrent neural networks: a survey , 1995, IEEE Trans. Neural Networks.

[8] Paul J. Werbos,et al. Backpropagation Through Time: What It Does and How to Do It , 1990, Proc. IEEE.

[9] Jeffrey L. Elman,et al. Finding Structure in Time , 1990, Cogn. Sci..