1,721,140 research outputs found
Skin detection for reducing false positive in Face Detection
No full textSkin detection is widely used in several applications ranging from tracking body parts to hand gesture analysis and face detection. In this chapter, we investigate and evaluate the usefulness of a skin detector to reduce the number of false positives found by an ensemble of face detectors.
The fusion of different face detectors permits on one hand to maximize the number of true positives found by the system, on the other (unfortunately) to increase also the number of false positives. To overcome this shortcoming difficulty, in this work we propose and test several filtering steps based firstly on skin detection, secondly on eye detection and when available on the depth map.
In this chapter, we investigate and evaluate an ensemble of approaches for skin detection based on different classifiers, color space and color constancy pre-processing. The proposed skin detection filtering step is first validated and compared with other state-of-the-art approaches on different skin datasets: a benchmark based on 25 videos of 8991 images with manually annotated pixel-level ground truth, the MCG-Skin benchmark dataset, the Pratheepan’s dataset and the UChile dbskin2 dataset. Then skin detection filtering step is evaluated for face detection purposes on two face datasets (one including both 2D and depth data). Experimental results confirm that our proposed approach obtains a very good performance. A MATLAB version of the filtering steps and the dataset used in this paper will be freely available from (https://www.dei.unipd.it/node/2357 + Pattern Recognition and Ensemble Classifiers)
Introduction to Local Binary Patterns: New Variants and Applications
No full textThis chapter provides an introduction to Local Binary Patterns (LBP) and important new variants. Some issues with LBP variants are discussed. A summary of the chapters on LBP is also presented
An ensemble of classifiers based on different texture descriptors for texture classification
No full textAbstractHere we propose a system that incorporates two different state-of-the-art classifiers (support vector machine and gaussian process classifier) and two different descriptors (multi local quinary patterns and multi local phase quantization with ternary coding) for texture classification.Both the tested descriptors are an ensemble of stand-alone descriptors obtained using different parameters setting (the same set is used in each dataset). For each stand-alone descriptor we train a different classifier, the set of scores of each classifier is normalized to mean equal to zero and standard deviation equal to one, then all the score sets are combined by the sum rule.Our experimental section shows that we succeed in building a high performance ensemble that works well on different datasets without any ad hoc parameters tuning. The fusion among the different systems permits to outperform SVM where the parameters and kernels are tuned separately in each dataset, while in the proposed ensemble the linear SVM, with the same parameter cost in all the datasets, is used
Deep semantic segmentation in skin detection
No full textDeep semantic segmentation is a task that identifies objects
and their boundaries in images, to do that a classification task is performed
at the pixel level to tag whether a pixel belongs to an object. In skin
detection, areas of images are classified as skin or non-skin regions. In
this work, we report a short survey of the recent literature covering the
task to help researchers in selecting the most suitable method for their
application and to expand the knowledge about the available datasets for
this topic. A compact empirical evaluation comparing recent models and
a new ensemble model is reported
Ensemble of both texture and color features for Reliable Person Re-identification
No full textHere a set of non trained person re-identification approaches is proposed for obtaining an high performance system, the proposed system has been tested on four datasets (CAVIAR4REID, IAS, BIWI and VIPeR). To reduce the risk of overfitting, for all the methods, the parameters have been kept constant across all datasets. The ensemble proposed in this work is based on different enhancement techniques, colorimetric spaces and state-of-the-art approaches. For the datasets where the depth map is available also a method based on skeleton detection, extracted from the depth map, belongs to the ensemble.
In our opinion, the proposed ensemble can be considered a general-purpose person re-identification system since all the parameters are not optimized separately in each dataset but are fixed.
The source code used for the approaches tested in this paper will be available at (https://www.dei.unipd.it/node/2357 + Pattern Recognition and Ensemble Classifiers)
An Empirical Study of Different Approaches for Protein Classification
Full text linkMany domains would benefit from reliable and efficient systems for automatic protein classification. An area of particular interest in recent studies on automatic protein classification is the exploration of new methods for extracting features from a protein that work well for specific problems. These methods, however, are not generalizable and have proven useful in only a few domains. Our goal is to evaluate several feature extraction approaches for representing proteins by testing them across multiple datasets. Different types of protein representations are evaluated: those starting from the position specific scoring matrix of the proteins (PSSM), those derived from the amino-acid sequence, two matrix representations, and features taken from the 3D tertiary structure of the protein. We also test new variants of proteins descriptors. We develop our system experimentally by comparing and combining different descriptors taken from the protein representations. Each descriptor is used to train a separate support vector machine (SVM), and the results are combined by sum rule. Some stand-alone descriptors work well on some datasets but not on others. Through fusion, the different descriptors provide a performance that works well across all tested datasets, in some cases performing better than the state-of-the-art
Convolutional neural networks for ATC classification
No full textAnatomical Therapeutic Chemical (ATC) classification of unknown compound has raised high significance for both drug development and basic research. The ATC system is a multi-label classification system proposed by the World Health Organization (WHO), which categorizes drugs into classes according to their therapeutic effects and characteristics. This system comprises five levels and includes several classes in each level; the first level includes 14 main overlapping classes. The ATC classification system simultaneously considers anatomical distribution, therapeutic effects, and chemical characteristics, the prediction for an unknown compound of its ATC classes is an essential problem, since such a prediction could be used to deduce not only a compound's possible active ingredients but also its therapeutic, pharmacological, and chemical properties. Nevertheless, the problem of automatic prediction is very challenging due to the high variability of the samples and the presence of overlapping among classes, resulting in multiple predictions and making machine learning extremely difficult
Heterogeneous bag of features for object/scene recognition
No full textIn this work we propose a method for object recognition based on a random selection of interest regions, heterogeneous set of texture descriptors and a bag-of-features approach based on several k- means clustering runs for obtaining different codebooks. The proposed system is not based on complex region detection as SIFT but on a simple exhaustive extraction of sub-windows of a given image. In the classification step an ensemble of random subspace of support vector machine (SVM) is used. The use of random subspace ensemble coupled to the principal component analysis for reducing the dimensionality of the descriptors permits to reduce the curse of dimensionality problem. In the experimental section we show that the combination of classifiers trained using different descriptors permits a consistent improvement of the performance of the stand alone approaches. The proposed system has been tested on four datasets: in the VOC2006 dataset, in a wide-used scene recognition dataset, in the well-known Caltech-256 Object Category Dataset and in a landmark dataset, obtaining remarkable results with respect to other state-of-the-art approaches. The MATLAB code of our system is publicly available
Texture descriptors for the generic pattern classification problem
No full textGood feature extraction methods are key in many pattern classification problems since the quality of pattern representations affects classification performance. Unfortunately, feature extraction is mostly problem dependent, with different descriptors typically working well with some problems but not with others. In this work, we propose a generalized framework that utilizes matrix representation for extracting features from patterns that can be effectively applied to very different classification problems. The idea is to adopt a two-dimensional representation of patterns by reshaping vectors into matrices so that powerful texture descriptors can be extracted. Since texture analysis is one of the most fundamental tasks used in computer vision, a number of high performing methods have been developed that have proven highly capable of extracting important information about the structural arrangement of pixels in an image (that is, in their relationships to each other and their environment). In this work, first, we propose some novel techniques for representing patterns in matrix form. Second, we extract a wide variety of texture descriptors from these matrices. Finally, the proposed approach is tested for generalizability across several well-known benchmark datasets that reflect a diversity of classification problems. Our experiments show that when different approaches for transforming a vector into a matrix are combined with several texture descriptors the resulting system works well on many different problems without requiring any ad-hoc optimization. Moreover, because texture-based and standard vector-based descriptors preserve different aspects of the information available in patterns, our experiments demonstrate that the combination of the two improves overall classification performance. The MATLAB code for our proposed system will be publicly available to other researchers for future comparisons
An empirical study on the matrix based protein representations and their combination with sequence based approaches
Full text linkMany domains have a stake in the development of reliable systems for automatic protein classification. Of particular interest in recent studies of automatic protein classification is the exploration of new methods for extracting features from a protein that enhance classification for specific problems. These methods have proven very useful in one or two domains, but they have failed to generalize well across several domains (i.e. classification problems). In this paper we evaluate several feature extraction approaches for representing proteins with the aim of sequence-based protein classification. Several protein representation are evaluated, those starting from: the position specific scoring matrix (PSSM) of the proteins; the amino-acid sequence; a matrix representation of the protein, of dimension (length of the protein)×20, obtained using the substitution matrices for representing each amino-acid as a vector. A valuable result is that a texture descriptor can be extracted from the PSSM protein representation which improve the performance of standard descriptors based on the PSSM representation. Experimentally we develop our systems by comparing several protein descriptors on nine different datasets. Each descriptor is used to train a support vector machine (SVM) or an ensemble of SVM. Although different stand-alone descriptors work well on some datasets (but not on others), we have discovered that fusion among classifiers trained using different descriptors obtains a good performance across all the tested datasets. Matlab code/Datasets used in the proposed paper is available at bias.csr.unibo.it\nanni\PSSM.rar
- …
