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.
Cross-domain matching is a challenging problem with several applications like face recognition across pose and resolution, heterogeneous face recognition, etc. Coupled dictionary learning has emerged as a powerful technique for addressing such problems. A novel approach based on aligning two orthogonal dictionaries constructed independently from the two domains is proposed in this work. Once the dictionaries...
In our earlier work, we have explored the sparse representation classification (SRC) for language recognition (LR) task. In those works, the orthogonal matching pursuit (OMP) algorithm was used for sparse coding. In place of l0-norm minimization in the OMP algorithm, one could also use ll-norm minimization based sparse coding such as the least absolute shrinkage and selection operator (LASSO). Though...
In this work, we explore the use of sparse features derived using a learned dictionary for language recognition (LR). These sparse features are referred to as s-vector and are derived by sparse coding of the commonly used low-dimensional i-vector based representation of speech utterances over the learned dictionary. The orthogonal matching pursuit (OMP), least absolute shrinkage and selection operator...
Sound is only the important way of human perception on the world, but also reflects some important characteristics of human behaviors under certain circumstances. This paper mainly presents the sound event recognition method using Mel-frequency Morlet wavelet subband (MFMWS) feature and sparse representation-based classifier (SRC) aiming at security monitoring applications. In order to evaluate its...
This paper presents a novel technique of image classification using BOVW model. The entire process first involves feature detection of images using FAST, the choice made in order to speed up the process of detection. Then comes the stage of feature extraction for which FREAK, a binary feature descriptor is employed. K-means clustering is then applied in order to make the bag of visual words. Every...
The problem that how to improve the market impact prediction performances of predictors that are trained based on stocks with few market news is studied in this preliminary work. We propose sentimental transfer learning to transfer the knowledge learned from news-rich stocks that are within the same sector to the news-poor stocks. News articles of both kinds of stocks are mapped into the same feature...
Driver's gaze direction is an indicator of driver state and plays a significantly role in driving safety. Traditional gaze zone estimation methods based on eye model have disadvantages due to the vulnerability under large head movement. Different from these methods, an appearance-based head pose-free eye gaze prediction method is proposed in this paper, for driver gaze zone estimation under free head...
Sentiment analysis is a process of identifying and categorizing opinions expressed in a piece of text. It classifies the text into positive, negative or neutral. Lexicon-based and Supervised Machine Learning-based are the two main approaches in sentiment analysis. Bag-of-words model is used to represent the text as a vector of independent words and machine learning algorithms are used for classification...
Extreme machine learning and its variants have shown good generalization performance and high leaning speed in many applications through its fast convergence. Despite the parallel and distributed ELM on MapReduce framework able to handle very large scale dataset for bigdata applications, the process of coping up with the rapidly updating data is a challenging one. Among the unified algorithms, the...
Sparse representation is a novel methodology that has off late received substantial attention for image classification and recognition. This paper presents a PCA-based dictionary building for sparse recognition. Recursive least square based auto-associative neural network model has been used for principal component extraction. Suggested network structure supports data compression along with principal...
Intrusion resilience is a protection strategy aimed at building systems that can continue to provide service during attacks. One approach to intrusion resilience is to continuously monitor a system's state and change its configuration to maintain service even while attacks are occurring. Intrusion detection, through both anomaly detection (for unknown attacks) and signature detection (for known attacks)...
The P300 event-related potential is often used as input signal for BCI control. BCI researchers often invest time in studies on stimulation and classification procedures that remain below results already achieved. Translational studies are sparse and also their results have only little impact on BCI research and development. Potential reasons for the lack of substantial translational research including...
Self-taught learning (STL) has become a promising paradigm to exploit unlabeled data for classification. The most commonly used approach to self-taught learning is sparse representation, in which it is assumed that each sample can be represented by a weighted linear combination of elements of a unlabeled dictionary. This paper proposes discriminative archetypal self-taught learning for the application...
Domain adaptation aims to deal with a kind of problem, in which the distribution of training scenarios and testing scenarios are different. Traditional solutions consider this problem in the point of the distribution matching. For the problem of domain adaptation of image classification, this paper proposes a new collaborative representation from the view of image representation. First, all source...
In the most of super-resolution reconstruction algorithms, high-resolution and low-resolution images are assumed in the same manifold space. However, due to distractions, this assumption is not suitable for the practical applications. This paper proposes a novel super-resolution reconstruction algorithm for face images. In order to consider the manifold inconsistency of high-resolution and low-resolution...
Because of the challenge of collecting labelled training data, zero-shot learning (ZSL) which transfers semantic knowledge represented by category attributes from seen classes to recognize unseen classes has received a lot of attention recently. Existing methods assume that the source attributes are completely correct in zero-shot learning. However, the source attributes in practice may contain noise...
Credit risk analysis seeks to determine whether a customer is likely to default on the financial obligation, which is a very important problem in finance. In this paper, we will present a machine learning framework to deal with this problem by formulating it as a binary classification problem. The framework consists of two parts: dictionary learning and classifier training. Firstly, we introduce a...
Low Light Level Images (LLLIs) are captured with exceptionally low brightness and low contrast, and cannot be enhanced satisfactorily with ordinary methods. In this paper, we propose a LLLI enhancement method using coupled dictionary learning. During the training stage, a pair of dictionaries and a linear mapping function are learned simultaneously. The dictionary pair aims to describe the raw LLLIs...
We propose mutually incoherent pose bases for action recognition in static image, each of which implicitly represents co-occurrence of poselets. First of all, action specific poselets are trained. To suppress the ambiguity of detection, we cluster poselet activations by the overlap of predicted torso bound of each poselet. Then pose feature of an action person can be extracted which is a vector composed...
In this paper, a fast method for single image super-resolution using dictionary learning is proposed. In this method, a local high resolution (HR) dictionary is constructed for every patch in the input image. To do this, the information from neighboring patches of the corresponding patch is used. Also, a low resolution (LR) dictionary consists of features obtained from patches of LR images in the...
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.