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.
Road pixel segmentation in airborne data is an important and challenging task. Recently, a sophisticated and robust approach based on superpixels and minimum cost paths has been published. In order to find out which of the numerous features are most essential, we propose a forward-search wrapper approach for feature selection which was tested with two different classifiers and with both generic and...
We propose a novel data-driven approach for automatically detecting and completing gaps in line drawings with a Convolutional Neural Network. In the case of existing inpainting approaches for natural images, masks indicating the missing regions are generally required as input. Here, we show that line drawings have enough structures that can be learned by the CNN to allow automatic detection and completion...
Many noninvasive continuous blood pressure measurements using photoplethysmography (PPG) are still inadequate in terms of accuracy and stability, which hinders the practical application of this method. This paper proposes a model based on ensemble method for BP estimation using PPG. A number of blood pressure calculation base-models is built on the same training data. These base-models are used to...
This paper presents a robust machine learning based computational solution for human detection. The proposed mechanism is specifically applicable for pose-variant situations in video frames. In order to address the pose variance problem, features are extracted using an improved variant of Histograms of Gradients (HoG) and local Binary Pattern features (LBP). The two feature sets are combined to form...
In the Gaussian mixture model based online writer identification system, the writer specific models are usually learned by adapting the universal background model. However, among all the possible adapting plans, which one performs best is still an unsolved problem, as well as the underlying principles. Towards finding the answer, this paper analyses all the combinations of the parameter adaptation...
BGM (background music) of a video plays an important role for making a video impressive. Although a large number of royalty-free music clips are available on the web, it is still difficult for amateur video creators to select appropriate music clips for their videos. In this paper, we propose a computational method for estimating the impression of a video from auditory and visual features of a video...
In this paper we propose an online multi-task learning algorithm for video concept detection. In particular, we extend the Efficient Lifelong Learning Algorithm (ELLA) in the following ways: a) we solve the objective function of ELLA using quadratic programming instead of solving the Lasso problem, b) we add a new label-based constraint that considers concept correlations, c) we use linear SVMs as...
This paper proposes a novel approach to voice conversion with non-parallel training data. The idea is to bridge between speakers by means of Phonetic PosteriorGrams (PPGs) obtained from a speaker-independent automatic speech recognition (SI-ASR) system. It is assumed that these PPGs can represent articulation of speech sounds in a speaker-normalized space and correspond to spoken content speaker-independently...
Most of the data in the field of social media has many features. Accordingly, one of the main challenges in this field is processing such high-dimensional data. Researchers are motivated to propose novel approaches in order to overcome this problem. One of the best solutions is extracting the effective information from data pool and discard unnecessary one. Feature selection is a known technique which...
This paper proposes a new method for recognizing both activities and gestures by using acceleration data collected on a smartwatch. While both activity recognition techniques and gesture recognition techniques employ acceleration data, these techniques are studied independently due to the large difference between the characteristics of activity sensor data and gesture sensor data. In this study, we...
Filtering pages about an entity (person, company, music band...) so that only interesting pages are kept is a real challenge. The interest can be qualified using criteria such as recency, novelty. In the last decade, we have seen classification systems trained to detect the interest for a document regarding an entity. For scalability reasons, it is not possible to consider a manual annotation of a...
Computers and Smartphone's becomes vital part of everyday life and hence use of internet becomes more and more. Due to internet, computers are becomes vulnerable of different kinds of security threats. Therefore it is required that we need to have efficient security method in order to avoid leakage of important data or misuse of data. This security method is called as Intrusion Detection System (IDS)...
Pedestrian detection attracts lots of attentions in the field of computer vision in recent years. It is difficult to handle data imbalance between positive and negative examples and easy-to-confused negative samples for pedestrian detection when training a single deep convolutional neural network (CNN) model. In this paper, we present a deep learning approach that combines two parallel deep CNN models...
In this paper, a computationally efficient voice conversion system has been designed in order to improve the performance in resource-limited scenarios. First, mixtures of Gaus-sians (MoGs) at fixed locations of Mel frequencies have been used to represent the spectrum of STRAIGHT compactly. Second, the key conditional distributions for prediction are approximated by building histograms of aligned features...
We investigate how overhead imagery can be integrated with non-image geographic data to learn appearance models for geographic objects with minimal user supervision. While multi-modal data integration has been successfully applied in other domains, such as multimedia analysis, significant opportunity remains for similar treatment of geographic data due to location being a simple yet powerful key for...
We address the problem of multimedia event detection from videos captured 'in the wild,' in particular the fusion of cues from multiple aspects of the video's content: detected objects, observed motion, audio signatures, etc. We employ score fusion, also known as late fusion, and propose a method that learns local weightings of the various base classifier scores which respect the performance differences...
A lot of research has been lately focusing on deep neural networks as an alternative to shallow ones. The added advantage among many, is the automated feature extraction of pattern from data. These models have been applied successfully to many tasks, including handwritten digit recognition, where they lead the state of the art performance. In this paper we apply a sparse deep belief network and a...
The advanced sensing and imaging technologies of today's digital camera systems provide the capability of monitoring traffic flows in a very large area. In order to provide continuous monitoring and prompt anomaly detection, an abstract-level autonomous anomaly detection model is developed that is able to detect various categories of abnormal vehicle events with unsupervised learning. The method is...
This paper proposes contour-based features for articulated pose estimation. Most of recent methods are designed using tree-structured models with appearance evaluation only within the region of each part. While these models allow us to speed up global optimization in localizing the whole parts, useful appearance cues between neighboring parts are missing. Our work focuses on how to evaluate parts...
We have studied the problem of classifying of surnames into the countries of origin using a collection of feature based learning algorithms. We have compiled a database of surnames and their countries of origin from publicly available databases as training data for the classifiers. We propose a feature selection algorithm which dynamically decides the most prominent feature of the names based on 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.