Obtained Feature Space Research Articles

A deep learning based interval type-2 fuzzy approach for image retrieval systems

Deep learning, that one of its key benefits is automated feature extraction, has become a principal solution for computer vision. This paper presents a Deep Type-2 Beta Fuzzy (DT2F) approach for Content-Based Image Retrieval (CBIR) systems. Firstly, the suggested architecture uses InceptionResNetv2 a pre-trained deep learning model on Image-Net data as a feature extractor. Secondly, the obtained feature space is fuzzified to handle the uncertainties associated with the extracted values of deep features. Thirdly, the reduction dimensionality step is efficiently applied using a Multi-Variational Auto-Encoder (MVAE) to reduce computational complexity and achieve better performance. Ultimately, we retrieve images using the nearest neighbors rule based on type-2 fuzzy similarity to having higher proximity sensitivity. Extensive experimentations were accomplished on various image-retrieving datasets of different scales the proposed system achieved an average precision of 97.15% exceeding other state-of-the-art methods over many systems on Corel datasets, which can open the door for several hybridization breakthroughs in the area of image retrieval.

Prediction of Task Performance From Physiological Features of Stress Resilience.

In this paper, we investigate the potential of generic physiological features of stress resilience in predicting air traffic control (ATC) candidates' performance in a highly-stressful low-fidelity ATC simulator scenario. Stress resilience is highlighted as an important occupational factor that influences the performance and well-being of air traffic control officers (ATCO). Poor stress management, besides the lack of skills, can be a direct cause of poor performance under stress, both in the selection process of ATCOs and later in the workplace. 40 ATC candidates, within the final stages of their selection process, underwent a stimulation paradigm for elicitation and assessment of various generic task-unrelated physiological features, related to resting heart rate variability (HRV) and respiratory sinus arrhythmia (RSA), acoustic startle response (ASR) and the physiological allostatic response, which are all recognized as relevant psychophysiological markers of stress resilience. The multimodal approach included analysis of electrocardiography, electromyography, electrodermal activity and respiration. We make advances in computational methodology for assessment of physiological features of stress resilience, and investigate the predictive power of the obtained feature space in a binary classification problem: prediction of high- vs. low-performance on the developed ATC simulator. Our novel approach yields a relatively high 78.16% classification accuracy. These results are discussed in the context of prior work, while considering study limitations and proposing directions for future work.

Adaptive Independent Subspace Analysis of Brain Magnetic Resonance Imaging Data

Methods for image registration, segmentation, and visualization of magnetic resonance imaging (MRI) data are used widely to help medical doctors in supporting diagnostics. The large amount and complexity of MRI data require looking for new methods that allow for efficient processing of this data. Here, we propose using the adaptive independent subspace analysis (AISA) method to discover meaningful electroencephalogram activity in the MRI scan data. The results of AISA (image subspaces) are analyzed using image texture analysis methods to calculate first order, gray-level co-occurrence matrix, gray-level size-zone matrix, gray-level run-length matrix, and neighboring gray-tone difference matrix features. The obtained feature space is mapped to the 2D space using the t-distributed stochastic neighbor embedding method. The classification results achieved using the k-nearest neighbor classifier with 10-fold cross-validation have achieved 94.7% of accuracy (and f-score of 0.9356) from the real autism spectrum disorder dataset.

A modified technique for smart textural feature selection to extract retinal regions of interest using image pre-processing

To increase the efficiency of the laser coagulation surgery the problem of the most accurate segmentation of fundus images is especially relevant. Fundus image segmentation is carried out with high accuracy using effective features and the minimum number of parameters for segmentation of a single image fragment. This paper describes a modified technique for smart textural feature selection to extract retinal regions of interest using image preprocessing algorithms. Preprocessing algorithms significantly influence the selected features which provide a minimum error of object recognition. In addition image preprocessing algorithms provide a more precise object selection. The informativeness of the obtained feature space is studied using discriminant data analysis. The best fragmentation block size segmentation and feature sets provides the necessary accuracy to identify regions of interest. Those regions are determined by the analysis of the following 4 classes of fundus images: exudates, thick, thin vessels and healthy areas. The advantages and disadvantages of the considered preprocessing algorithms were identified.

Nyström-based approximate kernel subspace learning

In this paper, we describe a method for the determination of a subspace of the feature space in kernel methods, which is suited to large-scale learning problems. Linear model learning in the obtained space corresponds to a nonlinear model learning process in the input space. Since the obtained feature space is determined only by exploiting properties of the training data, this approach can be used for generic nonlinear pattern recognition. That is, nonlinear data mapping can be considered to be a pre-processing step exploiting nonlinear relationships between the training data. Linear techniques can be subsequently applied in the new feature space and, thus, they can model nonlinear properties of the problem at hand. In order to appropriately address the inherent problem of kernel learning methods related to their time and memory complexities, we follow an approximate learning approach. We show that the method can lead to considerable operation speed gains and achieve very good performance. Experimental results verify our analysis.

Unsupervised kernel learning for abnormal events detection

In this paper, we propose a method to detect abnormal events using a novel unsupervised kernel learning algorithm. The key of our method is to learn a suitable feature space and the associated kernel function of the training samples. By considering the self-similarity property of training samples, we assume that the training samples will show the distinctly clustering property in the obtained feature space. Non-negative matrix factorization (NMF) is used to learn the feature space, and the support vector data description (SVDD) method is adopted to measure the clustering degree of instances in the feature space. We append the clustering constraints in the process of learning the feature space and use the bases produced by NMF as the projection matrix to construct the kernel function in SVDD. In other words, we incorporate the minimal enclosing sphere constraints within the NMF formulation. In the process of feature space learning, instances in the obtained feature space will be described better and better by an hypersphere. Our algorithm converges to a local optimal solution by applying an alternating optimization approach. Experimental results on three public datasets and the comparison to the state-of-the-art methods show that our method is effective in detecting and locating unknown abnormal behaviors.

Speaker Dependent and Independent Isolated Hindi Word Recognizer using Hidden Markov Model (HMM)

Hindi is very complex language with large number of phonemes and being used with various ascents in different regions in India. In this manuscript, speaker dependent and independent isolated Hindi word recognizers using the Hidden Markov Model (HMM) is implemented, under noisy environment. For this study, a set of 10 Hindi names has been chosen as a test set for which the training and testing is performed. The scheme instigated here implements the Mel Frequency Cepstral Coefficients (MFCC) in order to compute the acoustic features of the speech signal. Then, K-means algorithm is used for the codebook generation by performing clustering over the obtained feature space. Baum Welch algorithm is used for re-estimating the parameters, and finally for deciding the recognized Hindi word whose model likelihood is highest, Viterbi algorithm has been implemented; for the given HMM. This work resulted in successful recognition with 98.6% recognition rate for speaker dependent recognition, for total of 10 speakers (6 male, 4 female) and 97.5% for speaker independent isolated word recognizer for 10 speakers (male). General Terms Speech recognition, Isolated Hindi words.

Single Neuron Computation: From Dynamical System to Feature Detector

White noise methods are a powerful tool for characterizing the computation performed by neural systems. These methods allow one to identify the feature or features that a neural system extracts from a complex input and to determine how these features are combined to drive the system's spiking response. These methods have also been applied to characterize the input-output relations of single neurons driven by synaptic inputs, simulated by direct current injection. To interpret the results of white noise analysis of single neurons, we would like to understand how the obtained feature space of a single neuron maps onto the biophysical properties of the membrane, in particular, the dynamics of ion channels. Here, through analysis of a simple dynamical model neuron, we draw explicit connections between the output of a white noise analysis and the underlying dynamical system. We find that under certain assumptions, the form of the relevant features is well defined by the parameters of the dynamical system. Further, we show that under some conditions, the feature space is spanned by the spike-triggered average and its successive order time derivatives.

Segmenting a low-depth-of-field image using morphological filters and region merging

We propose a novel algorithm to partition an image with low depth-of-field (DOF) into focused object-of-interest (OOI) and defocused background. The proposed algorithm unfolds into three steps. In the first step, we transform the low-DOF image into an appropriate feature space, in which the spatial distribution of the high-frequency components is represented. This is conducted by computing higher order statistics (HOS) for all pixels in the low-DOF image. Next, the obtained feature space, which is called HOS map in this paper, is simplified by removing small dark holes and bright patches using a morphological filter by reconstruction. Finally, the OOI is extracted by applying region merging to the simplified image and by thresholding. Unlike the previous methods that rely on sharp details of OOI only, the proposed algorithm complements the limitation of them by using morphological filters, which also allows perfect preservation of the contour information. Compared with the previous methods, the proposed method yields more accurate segmentation results, supporting faster processing.

Information fusion for scene understanding from interferometric SAR data in urban environments

We present a framework for scene understanding from interferometric synthetic aperture radar data that is based on Bayesian machine learning and information extraction and fusion. A generic description of the data in terms of multiple models is automatically generated from the original signals. The obtained feature space is then mapped to user semantics representing urban scene elements in a supervised step. The procedure is applicable at multiple scales. We give examples of urban area classification and building recognition of Shuttle Radar Topography Mission data and of building reconstruction from submetric resolution Intermap data.