A Scene Recognition Method of Autonomous Developmental Network Based onMulti-sensor Fusion
YU Huijin1,2, FANG Yongchun1,2, WEI Zhixin1,2
1. Institute of Robotics and Automatic Information System, College of Artificial Intelligence, Nankai University, Tianjin 300071, China; 2. Tianjin Key Laboratory of Intelligent Robotics, Nankai University, Tianjin 300071, China
Abstract:Considering the low accuracy and poor adaptability of the existing scene recognition methods, the autonomous developmental neural network is applied to the robot scene recognition task, and two scene recognition methods combining the autonomous developmental network and multi-sensor fusion are proposed, namely, the robot scene recognition method based on weighted Bayesian fusion, and the scene recognition method based on data fusion of the same autonomous developmental network architecture, where the multi-sensor information is fused in the decision-making layer and the data layer, respectively, so as to improve the accuracy of scene recognition. Meanwhile, the autonomous developmental network improves the adaptability of the recognition method for various complex scenes. The proposed scene recognition method is tested and analyzed, which proves its effectiveness and practicability. In addition, the proposed method achieves better accuracy in scene recognition due to more efficient use of collected data through data fusion in the same network architecture.
[1] Khaleghi B, Khamis A, Karray F O, et al. Multisensor data fusion:A review of the state-of-the-art[J]. Information Fusion, 2013, 14(1):28-44. [2] Anitha R, Renuka S, Abudhahir A. Multi sensor data fusion algorithms for target tracking using multiple measurements[C]//IEEE International Conference on Computational Intelligence and Computing Research. Piscataway, USA:IEEE, 2013:1-4. [3] Zhang Q, Liu Y, Blum R S, et al. Sparse representation based multi-sensor image fusion for multi-focus and multi-modality images:A review[J]. Information Fusion, 2018, 40(40):57-75. [4] Wan G W, Yang X L, Cai R L, et al. Robust and precise vehicle localization based on multi-sensor fusion in diverse city scenes[C]//IEEE International Conference on Robotics and Automation. Piscataway, USA:IEEE, 2018:4670-4677. [5] Rakotovao T, Mottin J, Puschini D, et al. Multi-sensor fusion of occupancy grids based on integer arithmetic[C]//IEEE International Conference on Robotics and Automation. Piscataway, USA:IEEE, 2016:1854-1859. [6] Liang M, Yang B, Chen Y, et al. Multi-task multi-sensor fusion for 3D object detection[C]//IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, USA:IEEE, 2019:7337-7345. [7] Liu M Y, Chen R Z, Li D R, et al. Scene recognition for indoor localization using a multi-sensor fusion approach[J]. Sensors, 2017, 17(12). DOI:10.3390/s17122847. [8] Zhu H Y, Weibel J-B, Lu S J. Discriminative multi-modal feature fusion for RGBD indoor scene recognition[C]//IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, USA:IEEE, 2016:2969-2976. [9] Bijelic M, Gruber T, Mannan F, et al. Seeing through fog without seeing fog:Deep multimodal sensor fusion in unseen adverse weather[C]//IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway, USA:IEEE, 2020:11679-11689. [10] Weng J Y, McClelland J, Pentland A, et al. Autonomous mental development by robots and animals[J]. Science, 2001, 291(5504):599-600. [11] Weng J Y. Three theorems:Brain-like networks logically reason and optimally generalize[C]//International Joint Conference on Neural Networks. Piscataway, USA:IEEE, 2011:2983-2990. [12] 王东署,谭达佩,韦晓琴. 基于发育网络的人脸朝向识别研究[J].郑州大学学报(工学版),2017,58(5):23-27. Wang D S, Tan D P, Wei X Q. Face orientation recognition based on development network[J]. Journal of Zhengzhou University (Engineering Science), 2017, 38(5):23-27. [13] Castro-Garcia J L, Weng J Y. Emergent multilingual language acquisition using developmental networks[C]//International Joint Conference on Neural Networks. Piscataway, USA:IEEE, 2019:1-8. [14] Zheng Z J, Wu X, Weng J Y. Emergent neural Turing machine and its visual navigation[J]. Neural Networks, 2019, 110:116-130. [15] 钱夔,宋爱国,章华涛,等. 基于自主发育神经网络的机器人室内场景识别[J].机器人,2013,35(6):703-708. Qian K, Song A G, Zhang H T, et al. Robot indoor scenes recognition based on autonomous developmental neural network[J]. Robot, 2013, 35(6):703-708. [16] 余慧瑾,方勇纯. 基于改进型自主发育网络的机器人场景识别方法[J/OL].自动化学报,(2019-12-18)[2020-11-01].DOI:10.16383/j.aas.c180779. Yu H J, Fang Y C. A robot scene recognition method based on improved autonomous development network[J]. Acta Automatica Sinica, (2019-12-18)[2020-11-01]. DOI:10.16383/j.aas.c180779. [17] Weng J Y, Luciw M. Dually optimal neuronal layers:Lobe component analysis[J]. IEEE Transactions on Autonomous Mental Development, 2009, 1(1):68-85. [18] Keysers C, Perrett D I. Demystifying social cognition:A Hebbian perspective[J]. Trends in Cognitive Sciences, 2004, 8(11):501-507.