Abstract
This paper proposes a trace rule based self-organized map (SOM) model built upon a sparse 2-stage deep belief network (DBN). The combination of SOM and sparse DBN forms a hierarchical network where DBN serves as a V2 features detector while SOM layer learns to extract transformation invariant features guided by trace learning rule during training phase. The performance of our proposed method is evaluated by stimulus specific information (SSI) measuring and comparison with classic algorithms. It is demonstrated that trace rule based SOM model can generate more neurons with high SSI value which is beneficial to convey more useful and discriminative information for further object recognition.
Similar content being viewed by others
References
Bell AJ, Sejnowski TJ (1997) The independent components of natural scenes are edge filters. Vis Res 37:3327–3338
Bengio Y, Lamblin P, Popovici D, Larochelle H (2007) Greedy layer-wise training of deep networks. Advances in neural information processing systems
Bengior Y (2009) Learning deep architectures for AI. Found Trends Mach Learn 2(1):1–127
Coates A, Lee H, Ng AY (2011) An analysis of single layer networks in unsupervised feature learning. J Mach Learn Res 15:215–223
Geusebroek JM, Burghouts GJ, Smeulders AWM (2005) The Amsterdam library of object images. Int J Comput Vis 61(1):103–112
Hateren JHV, Schaaf AVD (1998) Independent component filters of natural images compared with simple cells in primary visual cortex. Proc Royal Soc Biol Sci 265(1394):359–66
Hinton GE, Salakhutdinov R (2006) Reducing the dimensiionality of data with neural networks. Science 313(5786):504–507
Jarrett K, Kavukcuoglu K, Ranzato MA, LeCun Y (2009) What is the best multi-stage architecture for object recognition?. In: ICCV, vol 30, pp 2146–2153
Ji N, Zhang J, Zhang C, Yin Q (2014) Enhancing performance of restricted boltzmann machines via log-sum regularization. Knowl-Based Syst 63:82–96
Kohonen T (1981) Self-organized formation of topologically correct feature maps. Biol Cybern 43(1):59–69
Krizhevsky A, Sutskever I, Hinton G (2012) Imagenet classification with deep convolutional neural networks. Adv Neural Inf Proces Syst 25(2):2012
LeCun Y, kavukcuoglu K, Farabet C (2010) Convolutional networks and applications in vision. IEEE Int Symp Circuits Syst 14(5):253–256
Lee H, Battle A, Raina R, Ng AY (2007) Efficient sparse coding algorithms. In: NIPS, pp 801–808
Lee H, Ekanadham C, Ng AY (2008) Sparse deep belief net model for visual area v2. Adv Neural Inf Proces Syst 20:873–880
Liu M, Zhang D (2016) Pairwise constraint-guided sparse learning for feature selection. IEEE Trans Cybern 46.1:298–310
Olshausen BA, Field DJ (1996) Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature 381(6583):607–609
Robinson L, Rolls ET (2015) Invariant visual object recognition: biologically plausible. Biol Cybern 109:505–535
Rolls ET (2012) Invariant visual object and face recognition: neural and computational bases, and a model. VisNet Front Comput Neurosci 6(35):1–70
Rolls ET, Treves A (2011) The neuronal encoding of information in the brain. Prog Neurobiol 95:448–490
Rolls ET, Treves A, Tovee MJ, Panzeri S (1997) Information in the neuronal representation of individual stimuli in the primate temporal visual cortex. J Comput Neurosci 4:309–333
Socher R, Huval B, Bhat B, Manning CD, Ng AY (2012) Convolutional-recursive deep learning for 3d object classification. In: NIPS, pp 665–673
Szegedy C, Liu W, Jia Y (2015) Going deeper with convolutions. In: CVPR, pp 1–9
Wallis G, Rolls ET (1996) A model of invariant object recognition in the visual system. Prog Neurobiol 51:167–194
Yang J, Yu K, Gong Y, Huang TS (2009) Linear spatial pyramid matching using sparse coding for image classification. In: CVPR, pp 1794–1801
Zhang J, Liang J, Zhao H (2013) Local energy pattern for texture classification using self-adaptive quantizatiion thresholds. IEEE Trans Image Process 22.1:31–42
Zhang J, Zhao H, Liang J (2013) Continuous rotation invariant local descriptors for texton dictionary-based texture classification. Comput Vis Image Underst 117.1:56–75
Zhang J, Liang J, Zhang C, Zhao H (2015) Scale invariant texture representation based on frequency decomposition and gradient orientation. Pattern Recogn Lett 51:57–62
Acknowledgments
This work was supported in part by the Project of National Natural Science Foundation of China (Grant Nos. 61076097, 61473257).
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
Assume that P(t) = 1/S, where S is the number of objects, we can derive the possible maximum SSI of a neuron by:
Rights and permissions
About this article
Cite this article
Cai, H., Wang, S., Liu, E. et al. Invariant object recognition based on combination of sparse DBN and SOM with temporal trace rule. Multimed Tools Appl 76, 12017–12034 (2017). https://doi.org/10.1007/s11042-016-3956-3
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-016-3956-3