The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
Deep Neural Networks have become increasingly popular due to their efficient realization in GPU hardware. Problems that were once considered computationally intensive to implement using Neural networks have now become possible due to the vast amount of flexibility and capability offered by the GPU and Deep networks combination. In this work, we attempt to improve the recognition rate for images, using...
A novel method for liveness detection of dorsal hand vein (DHV) based on AR model is proposed. Firstly, existing real DHV images are used to constitute a projection space based on modified principal component analysis (PCA). Unlike the previous works using the method of PCA, zero eigenvalues with their eigenvectors are used to constitute the projection space in this work. Secondly, test samples, including...
This paper deals with a complex symbol recognition process considering a large number of classes and only one training image per class. Furthermore, the response times of recognition system should be short and the interpretation of results must be easy. In this particular case, both statistical and structural methods are not the most suitable. A new composite descriptor and a similarity measure are...
In this paper, we present a feedforward training algorithm using Regularized Logistic Regression and Neural Networks to recognize handwritten objects. Furthermore, we intend to consider the effect of Gaussian noise in this procedure in order to examine the versatility of our approach. We might intend to transmit the image of our digits through an AWGN channel to a certain destination and then do the...
Aiming to the problem that semi-characters seriously affect character recognition accuracy in the process of automatic meter reading, we proposed a new method to recognize characters automatically which is based on improved BP neural network. First, we preprocess the image with morphology and locate the character area by combining projecting method and the characteristics of the gray transition, then...
The paper presents the design of three types of neural networks with different features, including traditional backpropagation networks, radial basis function networks and counterpropagation networks. Traditional backpropagation networks require very complex training process before being applied for classification or approximation. Radial basis function networks simplify the training process by the...
A novel system for 3D face recognition is presented in this paper. Firstly, we reduce the noise and move spikes from all the 3D faces. Secondly, we use Iterative Closet Point (ICP) to align all 3D face with the first person, and then for each face, we find the nose tip. Once the nose tip is successfully found, we crop a region, which is defined by a sphere radius of 100 mm centered at the nose tip...
Given a sequence of observable features of a linear dynamical system (LDS), we propose the problem of finding a representation of the LDS which is sparse in terms of a given dictionary of LDSs. Since LDSs do not belong to Euclidean space, traditional sparse coding techniques do not apply. We propose a probabilistic framework and an efficient MAP algorithm to learn this sparse code. Since dynamic textures...
Detecting the region of a license plate is the key component of the vehicle license plate recognition system. The spatial frequency characteristic in the license plate region usually varies more than in the background. In this paper, we propose a new approach for vehicle license plate localization using an optimal trade-off maximum average correlation height (TO-MACH) filter in the frequency domain...
This paper presents a combination of intelligent learning algorithm, the Support Vector Machine, and the recognition of star pattern in Celestial Navigation. Considering the star pattern recognition's character, noticing the advantages of SVM in learning competence, the paper proposes a solution to star pattern recognition with multi-kernel SVM. A multi-kernel algorithm bases on Genetic Programming...
Neural network plays a major role in the field of pattern recognition. For pattern recognition, a major drawback with traditional neural networks is that neural networks may easily be trapped in spurious states. Synergetic neural network (SNN) has been proposed in the literature to overcome this problem, however, when applying synergetic neural network on face recognition, the results are not satisfactory...
ATR (automatic target recognition), based on SAR (synthetic aperture radar) image, is crucial to the success of battlefield awareness and has become a very hot research topic. But many problems, which are caused by a great deal of missing, polluting, and superimposing of the signals, still generally exist in practical application, such as the low recognizing rate. How to recognize the useful signals...
We propose an appearance based eigenfeature regularization methodology for recognizing human activities. This regularization utilizes a 3-parameter based eigenmodel derived from the variances of within-class (activity) scatter matrix. Original eigenvalues are replaced by the model eigenvalues which facilitates in regularizing eigenfeatures corresponding to very small and zero eigenvalues and perform...
Facial expression recognition remains a challenging problem especially when the face is partially corrupted or occluded. We propose using a new classification method, termed Sparse Representation based Classification (SRC), to accurately recognize expressions under these conditions. A test vector is representable as a linear combination of vectors from its own class and so its representation as a...
This paper represents a currency recognition system using ensemble neural network (ENN). The individual neural networks (NN) in an ENN are trained via negative correlation learning. The object of using negative correlation learning (NCL) is to expertise the individuals in an ensemble on different parts or portion of input patterns. The available currencies in the market consist of new, old and noisy...
In this paper, we propose a neural network with a mental rotation function composed of two, rotation and comparison functions, and discuss its feasibility. The proposed network will output its result of judgment concerning their coincidence or non-coincidence, when two figures are input. If they coincide, the network will output the angle of difference between the two figures. The training for learning...
This paper proposed an enhanced algorithm of palmprint recognition. The 2D Gabor was done firstly to filter in the main direction and strengthen the primary line's information. Then we adopted wavelet transform to decompose the palmprint image, and extract the low frequency component. Two-Dimensional Principal Component Analysis(2DPCA) can avoid transforming from image matrix to 1D vector so as to...
Medical image recognition is crucial step of medical image processing and has become a very hot research topic. But many problems, which are caused by a great deal of missing, polluting, and superimposing of the signals, still generally exist in practical application, such as the low recognizing rate. How to recognize the useful signals from badly polluted images has become a difficult point in medical...
This paper proposes a method for constructing a discriminative rotation invariant object recognition system from the set of complex moments by using a multi-class boosting algorithm. Experimental results show that a large of number images can be discriminated accurately with only a small number of features. This basically means economy of computational effort in feature acquisition and also possibility...
In this work we investigate the performance of Advanced Correlation Filters (ACFs) in the automatic classification of partial shoeprints for use in forensic science. In particular, the Optimum Trade-off Synthetic Discriminant Function (OTSDF) filter is used to match low quality partial shoeprints. Experiments were conducted on a database of images of 100 different shoes available on the market. For...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.