Zhiqiang Zhuo

Wavelet transform and texture recognition based on spiking neural network for visual images

Abstract The functionalities of spiking neurons can be applied to deal with biological stimuli and explain complicated intelligent behaviors of the brain. The wavelet transforms are widely used in image feature extraction and image compression. Based on the principles from the visual system and wavelet theory, spiking neural networks with the ON/OFF neuron pathways inspired from the human visual system are proposed to perform the fast wavelet transform and the reconstruction for visual images. By this way we try to simulate how the human brain uses the volition-controlled method to extract useful image information. Furthermore, we decompose each texture sample with the established networks and calculate the normalized energy of the obtained sub-images at different scales. These energy values are used as features for texture classification. The simulation results show that the spiking neural network can extract the main information of images so that the images can be accurately classified using the information.

Competitive behaviors of a spiking neural network with spike timing dependent plasticity

Spike timing dependent plasticity (STDP) learning rule is one of hot topics in neurobiology since its been widely believed that synaptic plasticity mainly contribute to learning and memory in brain. Up to now, STDP has been observed in a wide variety of areas of brain, hippocampus, cortex and so on. Competition among synapses is an important behavior for this learning rule. In present study, we propose a single layer spiking neural network model using STDP learning rule in inhibitory synapses to investigate the competitive behavior. The experiments show that the synapses among neurons are both strengthened on the whole training process. Thus neurons inhibit the activities of one another, eventually the neuron with the highest input spike rate win the competition. We have found that the behavior is efficient when the differences of firing rates of input neurons without STDP are great than 5Hz, otherwise the winner neuron is random. In order to use the principle to artificial intelligent system, we use a mechanism of dynamic learning rate to let the neuron with the highest input to be selected by the competitive behavior as the winner. Therefore, a robust competitive spiking neural network is obtained.

Pedestrian identification based on fusion of multiple features and multiple classifiers

Abstract Finding a specific person from videos in surveillance systems is a challenging task. In the videos, different people cannot be the same in a whole body appearance. Based on this fact, this paper has proposed new methods based on fusion of textures, angle histograms and color moments to find a specific person. The human visual system can discriminate different objects quickly and efficiently. Inspired by on-center and off-center receptive fields in the visual system, a network model based on spiking neurons is proposed to extract texture features, and it has behaviors similar to Gabor filters. According to human body proportion, a person image is divided into three parts: head, torso and leg. Texture features of three parts are extracted by means of this network. Back propagation neural network, multi-class SVM and KNN are used as classifiers. For improving recognition rate, different fusion methods have been studied such as the fusion of texture features and other features in three body parts, and decision fusion using voting mechanism, probability combination etc. The experimental results for different methods are provided and the best fusion method is proposed. The technology of Compute Unified Device Architecture is applied in the experiments, which greatly reduces the running time for extraction of texture features.

A Large-scale Images Processing Model Based on Hadoop Platform

This paper presents a parallel processing model based on Hadoop platform for large-scale images processing, which aims to make use of the advantages of high reliability and high scalability of Hadoop distributed platform for distributed memory and distributed computing, so as to achieve the purpose of fast processing of large-scale images. The Hadoop streaming technology is used in the model. The main operations are written on shell script as the mapper of Hadoop streaming, then an assigned filelist is used as the Hadoop streamings input. The large numbers of image files are delivered to cluster computers for concurrent image processing. The model has been implemented using virtual machines. A set of experimental results and analysis are provided.

GPU Implementation of Spiking Neural Networks for Edge Detection

Spiking neural networks (SNN) are effective model inspired by neural networks in the brain. However, when networks increase in size towards the biological scale, it is time-consuming to simulate the networks using CPU programming. To solve this problem, Graphic Processing Units (GPU) provide a method to speed up the simulation. It is proposed and proved as a pertinent solution for implementation of large scale of neural networks. This paper presents a GPU implementation of SNN for edge detection. The approach is then compared with an equivalent implementation on an Intel Xeon CPU. The results show that the GPU approach provide about 37 times faster than the CPU implementation.

A feature extraction algorithm based on 2D complexity of gabor wavelets transform for facial expression recognition

Facial expression recognition is one of challenging topic in image processing. In this paper, a new feature extraction algorithm is proposed for facial expression recognition, in which Gabor filter is combined with 2D complexity for feature extraction. In order to obtain information of texture of expression in a static gray image, the image is transformed to sub images by Gabor wavelet after considerable pretreatment, and then the complexities of sub images are calculated. Fast Principle component analysis (Fast-Pca) is used to reduce the dimensionality of 2D complexity of the sub images and the effectiveness of characteristic vector is tested through an learning vector quantization (LVQ) classifier. The proposed feature extraction algorithm has been successfully applied to the Japanese Female Facial Expression (JAFFE) database with 213 frontal images corresponding 10 different subjects. The images are acquired under variable illumination. Experimental results show that the proposed algorithm obtains low-dimension of features compared with traditional method and expression recognition accuracy is improved.

Segmentation Based on Spiking Neural Network Using Color Edge Gradient for Extraction of Corridor Floor

In this paper, for the purpose of obstacle avoidance for blind men in the environment of indoor corridor, a corridor ground segmentation algorithm is proposed using image processing mechanism of the human visual system combined with the existing segmentation algorithms in robot visual navigation techniques. The segmentation algorithm is based on a spiking neural network. First, three color image gradient maps are generated utilizing a spiking neural network. The best gradient map is generated from three color components to extract the effective and useful image edges. Then threshold segmentation method is used to eliminate unwanted gradient to identify the boundary of floor. Finally, the corridor ground is extracted. The experimental results show that the algorithm works efficiently and the boundary of floor can be extracted accurately for corridor images with certain noise textured and nontextured. The algorithm has the practicality and robustness for identification of ground floor in blind navigation.

People recognition in multi-cameras using the visual color processing mechanism

Abstract In this paper a people recognition algorithm is proposed, in which the color processing mechanism is inspired by the biological visual system. The algorithm is constructed in two parts. In the first part a spiking neural network is proposed to extract the color features of the people images which are captured from videos, and a set of new features is generated by fusing the color features and color moments. In the second part, after a feature reduction, a Support Vector Machine is trained and then used to recognize a specific people. The algorithm has been successfully applied to recognize people in CASIA Database with a high recognition rate. In order to evaluate performance and analyze characteristics of people recognition algorithms in multi-camera scenes, Multi-Camera Video (MCV) dataset is made in this paper. It is used to evaluate and analyze the proposed method and a set of characteristics of the proposed algorithms are obtained. Experimental results demonstrate that the algorithm is comparable to state-of-the-art approaches in terms of accuracy and the direction for further improvement of the proposed algorithm is provided.

Finding Specific Person Using Spiking Neural Network Based on Texture Features

This paper has proposed a new texture-based method to find a specific person. A spike neural network is constructed to extract texture features. The network is constructed using integrate-and-fire neuron model, and it has behaviors similar to the Gabor filters. The person image is divided into three parts: head, torso and leg. Three parts of texture features are extracted by means of this network. BP neural network and multi-class Support Vector Machine are used as classifiers. CUDA technology is applied in the experiment, which greatly reduces the computation time.

Moving People Recognition Algorithm Based on the Visual Color Processing Mechanism

In this paper a target people recognition algorithm is proposed, in which the color processing mechanism is inspired by the biological visual system. The algorithm is constructed in two parts. In the first part a spiking neural network is proposed to extract the color features of the objects which are captured from videos. In the second part, after a feature reduction, a Support Vector Machine is used to fuse the color features and recognize the target. The algorithm has been successfully applied to recognize target people appeared in video sequence with a high recognition rate and suitable for generic recognition domain.