Histogram Construction Research Articles

Efficient learning with projected histograms

High dimensional learning is a perennial problem due to challenges posed by the “curse of dimensionality”; learning typically demands more computing resources as well as more training data. In differentially private (DP) settings, this is further exacerbated by noise that needs adding to each dimension to achieve the required privacy. In this paper, we present a surprisingly simple approach to address all of these concerns at once, based on histograms constructed on a low-dimensional random projection (RP) of the data. Our approach exploits RP to take advantage of hidden low-dimensional structures in the data, yielding both computational efficiency, and improved error convergence with respect to the sample size—whereby less training data suffice for learning. We also propose a variant for efficient differentially private (DP) classification that further exploits the data-oblivious nature of both the histogram construction and the RP based dimensionality reduction, resulting in an efficient management of the privacy budget. We present a detailed and rigorous theoretical analysis of generalisation of our algorithms in several settings, showing that our approach is able to exploit low-dimensional structure of the data, ameliorates the ill-effects of noise required for privacy, and has good generalisation under minimal conditions. We also corroborate our findings experimentally, and demonstrate that our algorithms achieve competitive classification accuracy in both non-private and private settings.

Adaptive feature selection and optimized multiple histogram construction for reversible data hiding

Reversible data hiding (RDH) algorithms have been extensively employed in the field of copyright protection and information dissemination. Among various RDH algorithms, the multiple histogram modification (MHM) algorithm has attracted significant attention because of its capability to generate high-quality marked images. In previous MHM methods, the prediction error histograms were mostly generated in a fixed way. Recently, clustering algorithms automatically classify prediction errors into multiple classes, which enhances the similarity among prediction errors within the same class. However, the design of features and the determination of clustering numbers are crucial in clustering algorithms. Traditional algorithms utilize the same features and fix the number of clusters (e.g., empirically generate 16 classes), which may limit the performance due to the lack of adaptivity. To address these limitations, an adaptive initial feature selection scheme and a clustering number optimization scheme based on the Fuzzy C-Means (FCM) clustering algorithm are proposed in this paper. The superiority of the proposed scheme over other state-of-the-art schemes is verified by experimental results.

Multicenter Multivendor Evaluation of Dose Volume Histogram Creation Consistencies for 8 Commercial Radiation Therapy Dosimetric Systems

PurposeTo evaluate dose volume histogram (DVH) construction differences across eight major commercial treatment planning systems (TPSs) and dose reporting systems (DRSs) for clinically treated plans of various anatomical sites and target sizes. Materials and MethodsDose files from ten selected clinically treated plans with a hypo-fractionation, stereotactic radiotherapy prescription or sharp dose gradients such as head and neck plans ranging from prescription doses of 18 Gy in 1 fraction to 70 Gy in 35 fractions, each calculated at 0.25 cm and 0.125 cm grid size, were created and anonymized in Eclipse TPS, and exported to seven other major TPS (Pinnacle, RayStation, Elements) and DRS (MIM, Mobius, ProKnow, and Velocity) systems for comparison. Dose-volume constraint points of clinical importance for each plan were collected from each evaluated system (D0.03cc[Gy], volume, and the mean dose were used for structures without specified constraints). Each reported constraint type and structure volume was normalized to the value from Eclipse for a pairwise comparison. A Wilcoxon Rank Sum test was used for statistical significance and a multivariable regression model was evaluated adjusting for plan, grid size, and distance to target center. ResultsFor all DVH points relative to Eclipse, all systems reported median values within 1.0% difference of each other, however they were all different from Eclipse. Considering mean values, Pinnacle, RayStation, and Elements averaged at 1.038, 1.046, 1.024 respectively, while MIM, Mobius, ProKnow, and Velocity reported 1.026, 1.050, 1.033 and 1.022 respectively relative to Eclipse. Smaller dose grid size improved agreement between the systems marginally without statistical significance. For structure volumes relative to Eclipse, larger differences are seen across all systems with a range in median values up to 3.0% difference and mean up to 10.1% difference. ConclusionLarge variations were observed between all systems. Eclipse generally reported, at statistically significant levels, lower values than all other evaluated systems. The non-significant change resulting from lowering the dose grid resolution indicates that this resolution may be less important than other aspects of calculating DVH curves, such as the 3D modeling of the structure.

A Line Segment Detector for Space Target Images Robust to Complex Illumination

The relative pose estimation of the space target is indispensable for on-orbit autonomous service missions. Line segment detection is an important step in pose estimation. The traditional line segment detectors show impressive performance under sufficient illumination, while it is easy to fail under complex illumination conditions where the illumination is too bright or too dark. We propose a robust line segment detector for space applications considering the complex illumination in space environments. An improved two-dimensional histogram construction strategy is used to optimize the Otsu method to improve the accuracy of anchor map extraction. To further improve line segment detection’s effect, we introduce an aggregation method that uses the angle difference between segments, the distance between endpoints, and the overlap degree of segments to filter the aggregation candidate segments and connect disjoint line segments that probably came from the same segment. We demonstrate the performance of the proposed line segment detector using a variety of images collected on a semiphysical simulation platform. The results show that our method has better performance than traditional line segment detectors including LSD, Linelet, etc., in terms of line detection precision.

Fast Guided Median Filter.

Faster computation of a weighted median (WM) filter is impeded by the construction of a weighted histogram for every local window of data. Since the calculated weights vary for each local window, it is difficult, using a sliding window approach, to construct the weighted histogram efficiently. In this paper, we propose a novel WM filter that overcomes the difficulty of histogram construction. Our proposed method achieves real-time processing for higher resolution images and can be applied to multidimensional, multichannel, and high precision data. The weight kernel used in our WM filter is the pointwise guided filter, which is derived from the guided filter. The use of kernels based on the guided filter avoids gradient reversal artifacts and shows a higher denoising performance than the Gaussian kernel based on the color/intensity distance. The core idea of the proposed method is a formulation that allows the use of histogram updates with a sliding window approach to find the weighted median. For high precision data we propose an algorithm based on a linked list that can reduce the memory requirements of storing histograms and the computational cost of updating them. We present implementations of the proposed method that are suitable for both CPU and GPU. Experimental results show that the proposed method indeed realizes faster computation than conventional WM filters and is capable of filtering multidimensional, multichannel, and high precision data. This is an approach which is difficult to achieve with conventional methods.

Fast and fully-automated histograms for large-scale data sets

G-Enum histograms are a new fast and fully automated method for irregular histogram construction. By framing histogram construction as a density estimation problem and its automation as a model selection task, these histograms leverage the Minimum Description Length principle (MDL) to derive two different model selection criteria. Several proven theoretical results about these criteria give insights about their asymptotic behaviour and are used to speed up their optimisation. These insights, combined to a greedy search heuristic, are used to construct histograms in linearithmic time rather than the polynomial time incurred by previous works. The capabilities of the proposed MDL density estimation method are illustrated with reference to other fully automated methods in the literature, both on synthetic and large real-world data sets.

A Fast Two-Stage Bilateral Filter Using Constant Time O(1) Histogram Generation.

Bilateral Filtering (BF) is an effective edge-preserving smoothing technique in image processing. However, an inherent problem of BF for image denoising is that it is challenging to differentiate image noise and details with the range kernel, thus often preserving both noise and edges in denoising. This letter proposes a novel Dual-Histogram BF (DHBF) method that exploits an edge-preserving noise-reduced guidance image to compute the range kernel, removing isolated noisy pixels for better denoising results. Furthermore, we approximate the spatial kernel using mean filtering based on column histogram construction to achieve constant-time filtering regardless of the kernel radius’ size and achieve better smoothing. Experimental results on multiple benchmark datasets for denoising show that the proposed DHBF outperforms other state-of-the-art BF methods.

A Bag-of-Words Based Feature Extraction Scheme for American Sign Language Number Recognition from Hand Gesture Images

Human Computer Interaction (HCI) focuses on the interaction between humans and machines. An extensive list of applications exists for hand gesture recognition techniques, major candidates for HCI. The list covers various fields, one of which is sign language recognition. In this field, however, high accuracy and robustness are both needed; both present a major challenge. In addition, feature extraction from hand gesture images is a tough task because of the many parameters associated with them. This paper proposes an approach based on a bag-of-words (BoW) model for automatic recognition of American Sign Language (ASL) numbers. In this method, the first step is to obtain the set of representative vocabularies by applying a K-means clustering algorithm to a few randomly chosen images. Next, the vocabularies are used as bin centers for BoW histogram construction. The proposed histograms are shown to provide distinguishable features for classification of ASL numbers. For the purpose of classification, the K-nearest neighbors (kNN) classifier is employed utilizing the BoW histogram bin frequencies as features. For validation, very large experiments are done on two large ASL number-recognition datasets; the proposed method shows superior performance in classifying the numbers, achieving an F1 score of 99.92% in the Kaggle ASL numbers dataset.

Equi-Depth Histogram Construction Methodology for Big Data Tools

In recent decades, countless data sources such as social media, machines, and networks are constantly pushing data into the digital world. The size of the data has been growing exponentially. To understand the statistical information of data query optimization, equi-depth histograms are essential. In this paper, we present approximate equi-depth histogram construction for big data using both Apache Pig Scripts and Java Web Interface interacting with Apache Hadoop. We use equi-depth histogram construction with quality guarantees for big data approaches and implement them with Apache Hadoop Map-Reduce and Apache Pig user-defined functions. We introduce a prototype implementation of the construction of the approximate equi-depth histogram from the Java Server Face page using Apache Hadoop jobs and the Hadoop Distributed Files System, and we evaluate these methods using the demonstration. We explain Apache Pig Scripts techniques to create equi-depth histograms using big data. The results indicate that our system provides the capability of writing multiple jobs using Apache Pig, and programmers can make use of the advantages of Apache Pig to create histograms and eliminate the complex implementation of Map-Reduce jobs.

Using positional sequence patterns to estimate the selectivity of SQL LIKE queries

Sequence patterns are frequently employed in many expert system applications in a wide range of domains from bioinformatics to smart homes and stock market analysis. Regular sequence patterns fail to express whether two consecutive items in a pattern are occurring right after each other in all pattern occurrences in an item database or not. Such a differentiation may be important for many intelligent system applications, for instance, to better address business questions like “should two frequently-bought-together items be located right next to each other on a retail store shelf, or is it ok to place them at some distance as long as they are in the same aisle?”. In this paper, we propose a novel type of sequence pattern, called “positional sequence patterns”, and illustrate its application on a special expert system, i.e., the query planner/optimizer of a database management system. Positional sequence patterns allow to accommodate extra information regarding whether a frequent ordered item pair always occurs next to each other without any gap in between in all pattern occurrences. Since positional sequence patterns are not considered by the existing sequence pattern mining algorithms, we also propose an algorithm to mine them. Next, we integrate the positional sequence patterns into the selectivity estimation component of the query optimizer as an expert system application. More specifically, in the knowledgebase of the query optimizer, a histogram-like structure of positional sequence patterns are created and stored. Then, during query optimization time, these histograms are utilized to infer the selectivity of flexible text queries that are enabled by the SQL LIKE operator. In particular, the proposed selectivity estimation method employs redundant pattern elimination based on pattern information content during histogram construction, and a partitioning-based matching scheme. The experimental results on a real dataset from DBLP show that the proposed approach outperforms the state of the art by around 20% improvement in error rates.

Fuzzy Image Processing Based Architecture for Contrast Enhancement in Diabetic Retinopathy Images

Diabetic retinopathy” is damage to retina denotes one of the problems of diabetes which is a significant reason for visual infirmity and blindness. A comprehensive and routine eye check is important to early detection and rapid treatment. This study proposes a hardware system that can enhance the contrast in the diabetic retinopathy eye fundus images as a first step in different eye disease diagnoses. The fuzzy histogram equalization technique is proposed to increases the local contrast of Diabetic Retinopathy Images. First, a histogram construction hardware architecture for different image processing purposes has been built then modified with fuzzy techniques to create fuzzy histogram equalization architecture, which is used to enhance the original images. Both architectures are designed using a finite-state machine (FSM) technique and programmed with VHDL programming language. The first one is implemented using two (Spartan 3E-XC3S500 and Xilinx Artix-7 XC7A100T) kits, while the second architecture is implemented using (Spartan 3E-XC3S500) kit. The system consists also of a modified video graphics array (VGA) port to display the input and resulted images with a proper resolution. All the hardware outputs are compared to that results produce from MatLab for verification and the resulted images are tested by PSNR, MSE, ENTROPY ,and AMBE

SUM-optimal histograms for approximate query processing

In this paper, we study the problem of the SUM query approximation with histograms. We define a new kind of histogram called the SUM-optimal histogram which can provide better estimation result for the SUM queries than the traditional equi-depth and V-optimal histograms. We propose three methods for the histogram construction. The first one is a dynamic programming method, and the other two are approximate methods. We use a greedy strategy to insert separators into a histogram and use the stochastic gradient descent method to improve the accuracy of separators. The experimental results indicate that our method can provide better estimations for the SUM queries than the equi-depth and V-optimal histograms.

Low Cost Histogram Implementation for Image Processing using FPGA

The solution for various applications of industrial programmed testing is image processing. In spite of this, if the contents of the image are not available in an appropriate format, even the most complex algorithms cannot extract the right information, via using a histogram, the image content can be easily processed. In this paper a histogram construction hardware that it can be able to create a histogram for different types of gray scale images of different sizes, is presented which gives accurately equivalent data as derived from a histogram plot using ISE simulator and MATLAB. By taking benefit of the high-level features of the Spartan3EXC3S500E FPGA architectures and Artix-7XC7A100T, the execution time for the proposed design is 156 times faster than MATLAB time for (16×16) pixel image size in XC3S500E and it is 6 in the worst case. Also, the execution time for the proposed design is 236 times faster than MATLAB time for (16×16) (16×16) pixel image size in XC7A100T and it is 8 in the worst case. The maximum frequency resulted and the maximum number of resources used in XC3S500E is 160.411 MHz, 45% respectively. While the maximum frequency resulted and the maximum number of resources used in XC7A100T is 216.685MHz, 59 %.

Machine learning templates for QCD factorization in the search for physics beyond the standard model

High-multiplicity all-hadronic final states are an important, but difficult final state for searching for physics beyond the Standard Model. A powerful search method is to look for large jets with accidental substructure due to multiple hard partons falling within a single jet. One way for estimating the background in this search is to exploit an approximate factorization in quantum chromodynamics whereby the jet mass distribution is determined only by its kinematic properties. Traditionally, this approach has been executed using histograms constructed in a background-rich region. We propose a new approach based on Generative Adversarial Networks (GANs). These neural network approaches are naturally unbinned and can be readily conditioned on multiple jet properties. In addition to using vanilla GANs for this purpose, a modification to the traditional WGAN approach has been investigated where weight clipping is replaced by drawing weights from a naturally compact set (in this case, the circle). Both the vanilla and modified WGAN approaches significantly outperform the histogram method, especially when modeling the dependence on features not used in the histogram construction. These results can be useful for enhancing the sensitivity of LHC searches to high-multiplicity final states involving many quarks and gluons and serve as a useful benchmark where GANs may have immediate benefit to the HEP community.

Compressed Image Hashing using Minimum Magnitude CSLBP

Image hashing allows compression, enhancement or other signal processing operations on digital images which are usually acceptable manipulations. Whereas, cryptographic hash functions are very sensitive to even single bit changes in image. Image hashing is a sum of important quality features in quantized form. In this paper, we proposed a novel image hashing algorithm for authentication which is more robust against various kind of attacks. In proposed approach, a short hash code is obtained by using minimum magnitude Center Symmetric Local Binary Pattern (CSLBP). The desirable discrimination power of image hash is maintained by modified Local Binary Pattern(LBP) based edge weight factor generated from gradient image. The proposed hashing method extracts texture features using the Center Symmetric Local Binary Pattern (CSLBP). The discrimination power of hashing is increased by weight factor during CSLBP histogram construction. The generated histogram is compressed to 1/4 of the original histogram by minimum magnitude CSLBP. The proposed method, has a twofold advantage, first is a small length and second is acceptable discrimination power. Experimental results are demonstrated by hamming distance, TPR, FPR and ROC curves. Therefore the proposed method successfully does a fair classification of content preserving and content changing images.

Emotion Recognition from Speech Using the Bag-of-Visual Words on Audio Segment Spectrograms

It is noteworthy nowadays that monitoring and understanding a human’s emotional state plays a key role in the current and forthcoming computational technologies. On the other hand, this monitoring and analysis should be as unobtrusive as possible, since in our era the digital world has been smoothly adopted in everyday life activities. In this framework and within the domain of assessing humans’ affective state during their educational training, the most popular way to go is to use sensory equipment that would allow their observing without involving any kind of direct contact. Thus, in this work, we focus on human emotion recognition from audio stimuli (i.e., human speech) using a novel approach based on a computer vision inspired methodology, namely the bag-of-visual words method, applied on several audio segment spectrograms. The latter are considered to be the visual representation of the considered audio segment and may be analyzed by exploiting well-known traditional computer vision techniques, such as construction of a visual vocabulary, extraction of speeded-up robust features (SURF) features, quantization into a set of visual words, and image histogram construction. As a last step, support vector machines (SVM) classifiers are trained based on the aforementioned information. Finally, to further generalize the herein proposed approach, we utilize publicly available datasets from several human languages to perform cross-language experiments, both in terms of actor-created and real-life ones.

Region Space Guided Transfer Function Design for Nonlinear Neural Network Augmented Image Visualization

Visualization provides an interactive investigation of details of interest and improves understanding the implicit information. There is a strong need today for the acquisition of high quality visualization result for various fields, such as biomedical or other scientific field. Quality of biomedical volume data is often impacted by partial effect, noisy, and bias seriously due to the CT (Computed Tomography) or MRI (Magnetic Resonance Imaging) devices, which may give rise to an extremely difficult task of specifying transfer function and thus generate poor visualized image. In this paper, firstly a nonlinear neural network based denoising in the preprocessing stage is provided to improve the quality of 3D volume data. Based on the improved data, a novel region space with depth based 2D histogram construction method is then proposed to identify boundaries between materials, which is helpful for designing the proper semiautomated transfer function. Finally, the volume rendering pipeline with ray-casting algorithm is implemented to visualize several biomedical datasets. The noise in the volume data is suppressed effectively and the boundary between materials can be differentiated clearly by the transfer function designed via the modified 2D histogram.

Performance evaluation of the three-point angular correlation function

Abstract In recent years, we have observed an increase in the diversity of the processors ecosystem. Different designs and architectures are being studied based on their performance and power characteristics. While using benchmarks for this purpose allows for reproducibility and easy understanding of the results, using real scientific applications allows researchers to realize the actual implications of each design on the overall performance of their codes. This paper analyzes the performance of different implementations of a three-point angular correlation function. This function is used in the study of Large-Scale Distribution of galaxies in a variety of computational platforms. The function is based on histogram construction and presents a large computational cost. This cost dramatically increases with the size of the datasets. We have considered two different GPUs, a set of x86 Intel machines (multi- and many-core), ARM chipsets, as well as an FPGA. We first study the best possible implementation for each platform in terms of time to solution. We then compare the power used by those platforms for a predefined number of datasets. Energy is one of the main constraints that computer architects are facing nowadays. The results will be used to evaluate the performance of this function considering those two targets – time and energy – for those platforms and to analyze the suitability of each of those platforms for this specific problem.

Differentially Private Event Histogram Publication on Sequences over Graphs

The big data era is coming with strong and ever-growing demands on analyzing personal information and footprints in the cyber world. To enable such analysis without privacy leak risk, differential privacy (DP) has been quickly rising in recent years, as the first practical privacy protection model with rigorous theoretical guarantee. This paper discusses how to publish differentially private histograms on events in time series domain, with sequences of personal events over graphs with events as edges. Such individual-generated sequences commonly appear in formalized industrial workflows, online game logs, and spatial-temporal trajectories. Directly publishing the statistics of sequences may compromise personal privacy. While existing DP mechanisms mainly target at normalized domains with fixed and aligned dimensions, our problem raises new challenges when the sequences could follow arbitrary paths on the graph. To tackle the problem, we reformulate the problem with a three-step framework, which 1) carefully truncates the original sequences, trading off errors introduced by the truncation with those introduced by the noise added to guarantee privacy, 2) decomposes the event graph into path sub-domains based on a group of event pivots, and 3) employs a deeply optimized tree-based histogram construction approach for each sub-domain to benefit with less noise addition. We present a careful analysis on our framework to support thorough optimizations over each step of the framework, and verify the huge improvements of our proposals over state-of-the-art solutions.

Optimization and Automation of the Construction of Smooth Free Energy Profiles.

An adaptive procedure is introduced to construct smooth analytical profiles of the free energy along a reaction coordinate using sampled data from multiple biased simulations. The procedure is based upon identifying problematic regions encountered in maximum likelihood estimators of the profile where there are statistically relevant discrepancies between the empirical and parametrized cumulative distribution functions and preferentially improving the construction of the parametric profile in these regions. The method is designed to produce continuous and smooth analytical fits that satisfy statistical goodness-of-fit tests with a minimum number of parameters. The accuracy of the profile obtained from the adaptive construction is compared by numerical computation to that of smooth interpolations based on an optimally chosen weighted histogram method for a solvated ion pair system and for an activated process for which the analytical form of the potential of mean force is available. In the model where the exact profile is known, the adaptive procedure is shown to reduce the integrated error relative to the optimal histogram construction by a factor of 3 or more in the typical case where the sampling is not extensive. It is demonstrated that the adaptive procedure can be used to produce statistically accurate smooth analytical representations of the free energy profile that can be evaluated with little computational effort and require little user input.