Robust Feature Set Research Articles

A robust statistical set of features for Amazigh handwritten characters

The main problem in the handwritten character recognition systems (HCR) is to describe each character by a set of features that can distinguish it from the other characters. Thus, in this paper, we propose a robust set of features extracted from isolated Amazigh characters based on decomposing the character image into zones and calculate the density and the total length of the histogram projection in each zone. In the experimental evaluation, we test the proposed set of features, to show its performance, with different classification algorithms on a large database of handwritten Amazigh characters. The obtained results give recognition rates that reach 99.03% which we presume good and satisfactory compared to other approaches and show that our proposed set of features is useful to describe the Amazigh characters.

A semi-automated system based on level sets and invariant spatial interrelation shape features for Caenorhabditis elegans phenotypes

Caenorhabditis elegans shares several molecular and physiological homologies with humans and thus plays a key role in studying biological processes. As a consequence, much progress has been made in automating the analysis of C. elegans. However, there is still a strong need to achieve more progress in automating the analysis of static images of adult worms. In this paper, a three-phase semi-automated system has been proposed. As a first phase, a novel segmentation framework, based on variational level sets and local pressure force function, has been introduced to handle effectively images corrupted with intensity inhomogeneity. Then, a set of robust invariant symbolic features for high-throughput screening of image-based C. elegans phenotypes are extracted. Finally, a classification model is applied to discriminate between the different subsets. The proposed system demonstrates its effectiveness in measuring morphological phenotypes in individual worms of C. elegans.

Clinical Features and Associated Likelihood of Primary Ciliary Dyskinesia in Children and Adolescents.

Primary ciliary dyskinesia (PCD), a genetically heterogeneous, recessive disorder of motile cilia, is associated with distinct clinical features. Diagnostic tests, including ultrastructural analysis of cilia, nasal nitric oxide measurements, and molecular testing for mutations in PCD genes, have inherent limitations. To define a statistically valid combination of systematically defined clinical features that strongly associates with PCD in children and adolescents. Investigators at seven North American sites in the Genetic Disorders of Mucociliary Clearance Consortium prospectively and systematically assessed individuals (aged 0-18 yr) referred due to high suspicion for PCD. The investigators defined specific clinical questions for the clinical report form based on expert opinion. Diagnostic testing was performed using standardized protocols and included nasal nitric oxide measurement, ciliary biopsy for ultrastructural analysis of cilia, and molecular genetic testing for PCD-associated genes. Final diagnoses were assigned as "definite PCD" (hallmark ultrastructural defects and/or two mutations in a PCD-associated gene), "probable/possible PCD" (no ultrastructural defect or genetic diagnosis, but compatible clinical features and nasal nitric oxide level in PCD range), and "other diagnosis or undefined." Criteria were developed to define early childhood clinical features on the basis of responses to multiple specific queries. Each defined feature was tested by logistic regression. Sensitivity and specificity analyses were conducted to define the most robust set of clinical features associated with PCD. From 534 participants 18 years of age and younger, 205 were identified as having "definite PCD" (including 164 with two mutations in a PCD-associated gene), 187 were categorized as "other diagnosis or undefined," and 142 were defined as having "probable/possible PCD." Participants with "definite PCD" were compared with the "other diagnosis or undefined" group. Four criteria-defined clinical features were statistically predictive of PCD: laterality defect; unexplained neonatal respiratory distress; early-onset, year-round nasal congestion; and early-onset, year-round wet cough (adjusted odds ratios of 7.7, 6.6, 3.4, and 3.1, respectively). The sensitivity and specificity based on the number of criteria-defined clinical features were four features, 0.21 and 0.99, respectively; three features, 0.50 and 0.96, respectively; and two features, 0.80 and 0.72, respectively. Systematically defined early clinical features could help identify children, including infants, likely to have PCD. Clinical trial registered with ClinicalTrials.gov (NCT00323167).

Robust multilingual Named Entity Recognition with shallow semi-supervised features

We present a multilingual Named Entity Recognition approach based on a robust and general set of features across languages and datasets. Our system combines shallow local information with clustering semi-supervised features induced on large amounts of unlabeled text. Understanding via empirical experimentation how to effectively combine various types of clustering features allows us to seamlessly export our system to other datasets and languages. The result is a simple but highly competitive system which obtains state of the art results across five languages and twelve datasets. The results are reported on standard shared task evaluation data such as CoNLL for English, Spanish and Dutch. Furthermore, and despite the lack of linguistically motivated features, we also report best results for languages such as Basque and German. In addition, we demonstrate that our method also obtains very competitive results even when the amount of supervised data is cut by half, alleviating the dependency on manually annotated data. Finally, the results show that our emphasis on clustering features is crucial to develop robust out-of-domain models. The system and models are freely available to facilitate its use and guarantee the reproducibility of results.

Smartphone-Based Patients' Activity Recognition by Using a Self-Learning Scheme for Medical Monitoring.

Smartphone based activity recognition has recently received remarkable attention in various applications of mobile health such as safety monitoring, fitness tracking, and disease prediction. To achieve more accurate and simplified medical monitoring, this paper proposes a self-learning scheme for patients' activity recognition, in which a patient only needs to carry an ordinary smartphone that contains common motion sensors. After the real-time data collection though this smartphone, we preprocess the data using coordinate system transformation to eliminate phone orientation influence. A set of robust and effective features are then extracted from the preprocessed data. Because a patient may inevitably perform various unpredictable activities that have no apriori knowledge in the training dataset, we propose a self-learning activity recognition scheme. The scheme determines whether there are apriori training samples and labeled categories in training pools that well match with unpredictable activity data. If not, it automatically assembles these unpredictable samples into different clusters and gives them new category labels. These clustered samples combined with the acquired new category labels are then merged into the training dataset to reinforce recognition ability of the self-learning model. In experiments, we evaluate our scheme using the data collected from two postoperative patient volunteers, including six labeled daily activities as the initial apriori categories in the training pool. Experimental results demonstrate that the proposed self-learning scheme for activity recognition works very well for most cases. When there exist several types of unseen activities without any apriori information, the accuracy reaches above 80% after the self-learning process converges.

MGRA: Motion Gesture Recognition via Accelerometer.

Accelerometers have been widely embedded in most current mobile devices, enabling easy and intuitive operations. This paper proposes a Motion Gesture Recognition system (MGRA) based on accelerometer data only, which is entirely implemented on mobile devices and can provide users with real-time interactions. A robust and unique feature set is enumerated through the time domain, the frequency domain and singular value decomposition analysis using our motion gesture set containing 11,110 traces. The best feature vector for classification is selected, taking both static and mobile scenarios into consideration. MGRA exploits support vector machine as the classifier with the best feature vector. Evaluations confirm that MGRA can accommodate a broad set of gesture variations within each class, including execution time, amplitude and non-gestural movement. Extensive evaluations confirm that MGRA achieves higher accuracy under both static and mobile scenarios and costs less computation time and energy on an LG Nexus 5 than previous methods.

Computer-Aided Diagnosis with Deep Learning Architecture: Applications to Breast Lesions in US Images and Pulmonary Nodules in CT Scans.

This paper performs a comprehensive study on the deep-learning-based computer-aided diagnosis (CADx) for the differential diagnosis of benign and malignant nodules/lesions by avoiding the potential errors caused by inaccurate image processing results (e.g., boundary segmentation), as well as the classification bias resulting from a less robust feature set, as involved in most conventional CADx algorithms. Specifically, the stacked denoising auto-encoder (SDAE) is exploited on the two CADx applications for the differentiation of breast ultrasound lesions and lung CT nodules. The SDAE architecture is well equipped with the automatic feature exploration mechanism and noise tolerance advantage, and hence may be suitable to deal with the intrinsically noisy property of medical image data from various imaging modalities. To show the outperformance of SDAE-based CADx over the conventional scheme, two latest conventional CADx algorithms are implemented for comparison. 10 times of 10-fold cross-validations are conducted to illustrate the efficacy of the SDAE-based CADx algorithm. The experimental results show the significant performance boost by the SDAE-based CADx algorithm over the two conventional methods, suggesting that deep learning techniques can potentially change the design paradigm of the CADx systems without the need of explicit design and selection of problem-oriented features.

Image retrieval based on exponent moments descriptor and localized angular phase histogram

Multiple feature extraction and combination is one of the most important issues in the content-based image retrieval (CBIR). In this paper, we propose a new content-based image retrieval method based on an efficient combination of shape and texture features. As its shape features, exponent moments descriptor (EMD), which has many desirable properties such as expression efficiency, robustness to noise, geometric invariance, fast computation etc., is adopted in RGB color space. As its texture features, localized angular phase histogram (LAPH) of the intensity component, which is robust to illumination, scaling, and image blurring, is used in hue saturation intensity (HSI) color space. The combination of above shape and texture information provides a robust feature set for color image retrieval. Experimental results on well known databases show significant improvements in retrieval rates using the proposed method compared with some current state-of-the-art approaches.

Towards a Small Set of Robust Acoustic Features for Emotion Recognition: Challenges

The search of a small acoustic feature set for emotion recognition faces three main challenges. Such a feature set must be robust to large diversity of contexts in real-life applications; model parameters must also be optimized for reduced subsets; finally, the result of feature selection must be evaluated in cross-corpus condition. The goal of the present study is to select a consensual set of acoustic features for valence recognition using classification and non-classification based feature ranking and cross-corpus experiments, and to optimize emotional models simultaneously. Five realistic corpora are used in this study: three of them were collected in the framework of the French project on robotics ROMEO, one is a game corpus (JEMO) and one is the well-known AIBO corpus. Combinations of features found with non-classification based methods (information gain and Gaussian mixture models with Bhattacharyya distance) through multi-corpora experiments are tested under cross-corpus conditions, simultaneously with SVM parameters optimization. Reducing the number of features goes in pair with optimizing model parameters. Experiments carried on randomly selected features from two acoustic feature sets show that a feature space reduction is needed to avoid over-fitting. Since a Grid search tends to find non-standard values with small feature sets, the authors propose a multi-corpus optimization method based on different corpora and acoustic feature subsets which ensures more stability. The results show that acoustic families selected with both feature ranking methods are not relevant in cross-corpus experiments. Promising results have been obtained with a small set of 24 voiced cepstral coefficients while this family was ranked in the 2nd and 5th positions with both ranking methods. The proposed optimization method is more robust than the usual Grid search for cross-corpus experiments with small feature sets.

HOG와 칼만필터를 이용한 다중 표적 추적에 관한 연구

Detecting human in images is a challenging task owing to their variable appearance and the wide range of poses the they can adopt. The first need is a robust feature set that allows the human form to be discriminated cleanly, even in cluttered background under difficult illumination. A large number of vision application rely on matching keypoints across images. These days, the deployment of vision algorithms on smart phones and embedded device with low memory and computation complexity has even upped the ante: the goal is to make descriptors faster compute, more compact while remaining robust scale, rotation and noise. In this paper we focus on improving the speed of pedestrian(walking person) detection using Histogram of Oriented Gradient(HOG) descriptors provide excellent performance and tracking using kalman filter.

Computer vision for image-based transcriptomics

Single-cell transcriptomics has recently emerged as one of the most promising tools for understanding the diversity of the transcriptome among single cells. Image-based transcriptomics is unique compared to other methods as it does not require conversion of RNA to cDNA prior to signal amplification and transcript quantification. Thus, its efficiency in transcript detection is unmatched by other methods. In addition, image-based transcriptomics allows the study of the spatial organization of the transcriptome in single cells at single-molecule, and, when combined with superresolution microscopy, nanometer resolution. However, in order to unlock the full power of image-based transcriptomics, robust computer vision of single molecules and cells is required. Here, we shortly discuss the setup of the experimental pipeline for image-based transcriptomics, and then describe in detail the algorithms that we developed to extract, at high-throughput, robust multivariate feature sets of transcript molecule abundance, localization and patterning in tens of thousands of single cells across the transcriptome. These computer vision algorithms and pipelines can be downloaded from: https://github.com/pelkmanslab/ImageBasedTranscriptomics.

Improving cross-modal face recognition using polarimetric imaging.

We investigate the performance of polarimetric imaging in the long-wave infrared (LWIR) spectrum for cross-modal face recognition. For this work, polarimetric imagery is generated as stacks of three components: the conventional thermal intensity image (referred to as S0), and the two Stokes images, S1 and S2, which contain combinations of different polarizations. The proposed face recognition algorithm extracts and combines local gradient magnitude and orientation information from S0, S1, and S2 to generate a robust feature set that is well-suited for cross-modal face recognition. Initial results show that polarimetric LWIR-to-visible face recognition achieves an 18% increase in Rank-1 identification rate compared to conventional LWIR-to-visible face recognition. We conclude that a substantial improvement in automatic face recognition performance can be achieved by exploiting the polarization-state of radiance, as compared to using conventional thermal imagery.

Robust Spectral Features for Automatic Speaker Recognition in Mismatch Condition

Abstract The widespread use of automatic speaker recognition technology in real world applications demands for robustness against various realistic conditions. In this paper, a robust spectral feature set, called NDSF (Normalized Dynamic Spectral Features) is proposed for automatic speaker recognition in mismatch condition. Magnitude spectral subtraction is performed on spectral features for compensation against additive noise. A spectral domain modification is further performed using time-difference approach followed by Gaussianization Non-linearity. Histogram normalization is applied to these dynamic spectral features, to compensate the effect of channel mismatch and some non-linear effects introduced due to handset transducers. Feature extraction using proposed features is carried out for a text independent automatic speaker recognition (identification) system. The performance of proposed feature set is compared with conventional cepstral features like (mel-frequency cepstral coefficients and linear prediction cepstral coefficients), for acoustic mismatch condition caused by use of different sensors. Studies are performed on two databases: A multi-variability speaker recognition (MVSR) developed by IIT-Guwahati and Multi-speaker continuous (Hindi) speech database (By Department of Information Technology, Government of India). From experimental analysis, it is observed that, spectral domain dynamic features enhance the robustness by reducing additive noise and channel effects caused by sensor mismatch. The proposed NDSF features are found to be more robust than cepstral features for both datasets.

A Novel Hierarchical Technique for Offline Handwritten Gurmukhi Character Recognition

The increasing need of a handwritten character recognition system in the Indian offices such as banks, post offices and so forth, has made it an imperative field of research. In present paper, Authors have presented a novel hierarchical technique for isolated offline handwritten Gurmukhi character recognition. A robust feature set of 105 feature elements is proposed under this work for recognition of offline handwritten Gurmukhi characters using four types of topological features, namely, horizontally peak extent features, vertically peak extent features, diagonal features, and centroid features. For classification Support Vector Machines (SVMs) classifier has been used in this work. SVMs classifier has been considered with four different kernels, namely, linear kernel, polynomial kernel, RBF kernel and sigmoid kernel. For training and testing of a classifier, we have used 3,500 samples of isolated offline handwritten Gurmukhi characters written by one hundred different writers. Maximum recognition accuracy of 91.80 % have been achieved with proposed technique, while using PCA feature set and SVM with a linear kernel classifier.

English

This paper proposes a method of Content based image retrieval (CBIR) using color sketches, which is one of the most popular, rising research areas of the Digital image processing. The goal of CBIR is to extract visual content of an image automatically, like color, texture, or shape. The proposed method is based on free hand color sketch of image, which uses color and texture features. In this color features are extracted using HSV (Hue, Saturation and Value) color space. The HSV color space is quantified in non-equal intervals, a one dimensional feature vector is constructed and it is represented by cumulative histogram. Texture feature extraction is obtained by using gray-level co-occurrence matrix (GLCM). The combination of the color and texture features of an image provides a robust feature set for image retrieval. Euclidean distance of color and texture is used in retrieving the similar images. The image retrieval experiment indicates that the use of color and texture features in image retrieval using color sketches has obvious advantages. This paper will be very helpful in crime prevention. The proposed method has more retrieval rate than image retrieval using sketches.

Surface myoelectric signal classification using the AR-GARCH model

In myoelectric prostheses design, it is normally assumed that the necessary control information can be extracted from the surface myoelectric signals. In the pattern classification paradigm for controlling myoelectric prosthesis, the autoregressive (AR) model coefficients are generally considered an efficient and robust feature set. However, no formal statistical methodologies or tests are reported in the literature to analyze and model the myoelectric signal as an AR process. We analyzed the myoelectric signal as a stochastic time-series and found that the signal is heteroscedastic, i.e., the AR modeling residuals exhibit a time-varying variance. Heteroscedasticity is a major concern in statistical modeling because it can invalidate statistical tests of significance which may assume that the modeling errors are uncorrelated and that the error variances do not vary with the effects being modeled. We subsequently proposed to model the myoelectric signal as an autoregressive-generalized autoregressive conditional heteroscedastic (AR-GARCH) process and used the model parameters as a feature set for signal classification. Multiple statistical tests including the Ljung–Box Q-test, Engle's test for heteroscedasticity, the Kolmogorov–Smirnov test and the goodness of fit test were performed to show the validity of the proposed model. Our experimental results show that the proposed AR-GARCH model coefficients, when used as a feature set in two different classification schemes, significantly outperformed (p<.01) the conventional AR model coefficients.

MRI of Human Brain Images Classification Based on Dirichlet Laplacian

In [1], Dirichlet Laplacian (DL) was used to develop a three generic and robust set of features that are size-, translation-, and rotation-invariant that showed to be tolerant to boundary noise and deformation. The prescribed technique was applied to develop an efficient algorithm to classify the magnetic resonance images (MRI) and distinguish between the normal and abnormal images.

Visual voice activity detection via chaos based lip motion measure robust under illumination changes

In this paper, a vision based voice activity detection (VVAD) algorithm is proposed using chaos theory. In conventional VVAD algorithm, the movement measure of lip region is found by applying an optical flow algorithm to detect the visual speech frame using a motion based energy feature set. However, since motion based feature is unstable under illumination changes, a new form of robust feature set is desirable. It is propositioned that contextual changes such as lip opening or closing motion during speech utterances under illumination variation can be observed as chaos-like and the resultant complex fractal trajectories in phase space can be observed. The fractality is measured in phase space from two sequential video input frames and subsequently any visual speech frames are robustly detected. Representative experiments are performed in image sequence containing a driver scene undergoing illumination fluctuations in moving vehicle environment. Experimental results indicate that a substantial improvement is obtained in terms of achieving significantly lower false alarm rate over the conventional method.

PSO-Based Optimal Selection of Zernike Moments for Target Discrimination in High-Resolution SAR Imagery

Target discrimination is the key step of automatic target detection in synthetic aperture radar (SAR) images. In this paper, a new algorithm, effective and robust feature sets for target discrimination in high resolution SAR images has been proposed. Two main steps in target discrimination of SAR images have been developed, the feature extraction based on Zernike moments (ZMs) having linear transformation invariance properties and the PSO based feature selection to select the optimal feature subset of Zernike moments for decreasing computational complexity of feature extraction step. The input regions of interest (ROIs) have been segmented and passed to a number of preprocessing stages such as histogram equalization, position and size normalization. Two groups of Zernike moments (shape and margin (intensity) characteristic) have been extracted from the preprocessed images and they have been applied to the feature selection step. Each group includes 34 moments with different orders and iterations. The selected moments have been applied to a SVM classifier. The proposed scheme has been tested on the MSTAR database. The Receiver Operational Characteristics (ROC) curve and the performance of proposed method using some measured data have been analyzed. Experimental results demonstrate the efficiency of the proposed approach in target discrimination of SAR imagery.

Effective technique for the recognition of offline Arabic handwritten words using hidden Markov models

In this paper, we present a novel segmentation-free Arabic handwriting recognition system based on hidden Markov model (HMM). Two main contributions are introduced: a new technique for dividing the image into nonuniform horizontal segments to extract the features and a new technique for solving the problems of the skewing of characters by fusing multiple HMMs. Moreover, two enhancements are introduced: the pre-processing method and feature extraction using concavity space. The proposed system first pre-processes the input image by setting the thickness of the input word to three pixels and fixing the spacing between the different parts of the word. The input image is divided into constant number of nonuniform horizontal segments depending on the distribution of the foreground pixels. A set of robust features representing the gradient of the foreground pixels is extracted using sliding windows. The input image is decomposed into several images representing the vertical, horizontal, left diagonal and right diagonal edges in the image. A set of robust features representing the densities of the foreground pixels in the various edge images is extracted using sliding windows. The proposed system builds character HMM models and learns word HMM models using embedded training. Besides the vertical sliding window, two slanted sliding windows are used to extract the features. Three different HMMs are used: one for the vertical sliding window and two for the slanted windows. A fusion scheme is used to combine the three HMMs. The proposed system is very promising and outperforms all the other Arabic handwriting recognition systems reported in the literature.