A method of robot indoor scene recognition based on autonomous developmental neural network is proposed. 3-layer adaptive developmental neural network is used to build the brain-mind model. In developing phase, top-k competition is utilized to simulate the lateral inhibition of neurons, and the winner updates the synapse weight vector with the lobe component analysis (LCA) algorithm. The strengthened neurons can get thinking results according to current environment information, and indoor scenes can be recognized autonomously by mobile robots. Through human-like thinking, the learning results are stored as "knowledge", and the thinking results are derived from experience. Experimental results show that the model of autonomous developmental neural network proposed as the carrier of "knowledge", fully meets the need of indoor scenes recognition task, and realizes the autonomous learning, understanding and growth of robots based on vision.
[1] Albus J S. A model of computation and representation in the brain[J]. Information Sciences, 2010, 180(9): 1519-1554. [2] George D, Hawkins J. Towards a mathematical theory of cortical micro-circuits[J]. PLoS Computational Biology, 2009, 5(10): 1-26.[3] Tenenbaum J B, Griffiths T L, Kemp C. Theory-based Bayesian models of inductive learning and reasoning[J]. Trends in Cognitive Sciences, 2006, 10(7): 309-318. [4] Weng J. Three theorems: Brain-like networks logically reason and optimally generalize[C]//Proceeding of International Joint Conference on Neural Networks. Piscataway, USA: IEEE, 2011: 2983-2990.[5] Cox B R, Krichmar J L. Neuromodulation as a robot controller: A brain-inspired strategy for controlling autonomous robots[J]. IEEE Robotics and Automation Magazine, 2009, 16(3): 72-80. [6] Almassy N, Edelman G M, Sporns O. Behavioral constraints in the development of neuronal properties: A cortical model embedded in a real-world device[J]. Cerebral Cortex, 1998, 8(4): 346-361. [7] Solgi M, Weng J. Developmental stereo: Emergence of disparity preference in models of the visual cortex[J]. IEEE Transactions on Autonomous Mental Development, 2009, 1(4): 238-252. [8] 李桂芝, 安成万, 等.基于场景识别的移动机器人定位方法研究[J].机器人, 2005, 27(2):123-127. Li G Z, An C W, Yang G S, et al. Scene recognition for mobile robot localization[J]. Robot, 2005, 27(2): 123-127.[9] Vogel J, Schiele B. Natural scene retrieval based a semantic modeling step[C]//International Conference on Image and Video Retrieval. Berlin, Germany: Springer-Verlag, 2004: 207-215.[10] 张骏, 高隽, 谢昭, 等.基于统计分析Boosting的复杂场景目标识别方法研究[J].仪器仪表学报, 2010, 31(8):1788-1794. Zhang J, Gao J, Xie Z, et al. Object recognition in complex scenes based on statistical Boosting[J]. Chinese Journal of Scientific Instrument2010, 31(8): 1788-1794.[11] 高颖, 陈东岳, 张立明.一种带有实时视觉特征学习的自主发育机器人探索[J].复旦学报:自然科学版, 2005, 44(6):964-970. Gao Y, Chen D Y, Zhang L M. An exploration of autonomous developing Robot with Real time vision Learning[J]. Journal of Fudan University: Natural Science, 2005, 44(6): 964-970.[12] 陈洋, 张道辉, 赵新刚, 等.基于IHDR自主学习框架的无人机3维路径规划[J].机器人, 2012, 34(5):513-518. Chen Y, Zhang D G, Zhao X G, et al. UAV 3D path planning based on IHDR autonoumous-learning-framwork[J]. Robot, 2012, 34(5): 513-518.[13] Weng J, Luciw M. Brain-like emergent spatial processing[J]. IEEE Transactions on Autonomous Mental Development, 2012, 4(2): 161-185. [14] Weng J, Luciw M. Dually optimal neuronal layers: Lobe component analysis[J]. IEEE Transactions on Autonomous Mental Development, 2009, 1(1): 68-85. [15] Qian K, Song A G, Bao J T, et al. Small teleoperated robot for nuclear radiation and chemical leak detection[J]. International Journal of Advanced Robotic Systems, 2012, 9: No.70.[16] Luciw M, Weng J. Where-what network 3: Developmental topdown attention for multiple foregrounds and complex backgrounds[C]//IEEE International Joint Conference on Neural Networks. Piscataway, USA: IEEE, 2010: 4233-4240.[17] Lin K P, Chen M S. On the design and analysis of the privacypreserving SVM classifier[J]. IEEE Transactions on Knowledge and Data Engineering, 2011, 23(11): 1704-1717. [18] Ren Y, Bai G C. Determination of optimal SVM parameters by using GA/PSO[J]. Journal of Computers, 2010, 5(8): 1160-1168.)
