Showing papers in "Pattern Recognition Letters in 2017"

TL;DR: It appears that a system is in place to assist clinicians to diagnose seizures accurately in less time as the proposed model achieves perfect 100% classification sensitivity and is found to be outperforming all existing models in terms of classification sensitivity (CSE).

TL;DR: The first machine learning method for ground truthing a video is proposed, based on a multi-resolution convolutional neural network with a cascaded architecture, for segmenting foreground moving objects pictured in surveillance videos.

TL;DR: An automated method is proposed to easily differentiate between cancerous and non-cancerous Magnetic Resonance Imaging (MRI) of the brain and can be used to identify the tumor more accurately in less processing time as compared to existing methods.

TL;DR: This work proposes the meta-heuristic approach assisted segmentation and analysis of glioma from brain MRI dataset based on tri-level thresholding and level set segmentation, which achieved better values of Jaccard index, dice co-efficient, precision, sensitivity, specificity and accuracy.

TL;DR: This work has proposed a classification methodology to classify Focal and Non Focal EEG and found that NNge classifier gave the highest accuracy of 98%, sensitivity of 100% and specificity of 96%, which is the highest comparing to other methods in the literature.

TL;DR: Results demonstrate that DM and RM are very promising modalities for vehicle detection, and experiments show that the proposed fusion strategy achieves higher accuracy compared to each modality alone in all the levels of increasing difficulty in KITTI object detection dataset.

TL;DR: A novel multi-sensor fusion framework for Sign Language Recognition (SLR) using Coupled Hidden Markov Model (CHMM), which provides interaction in state-space instead of observation states as used in classical HMM that fails to model correlation between inter-modal dependencies.

TL;DR: This paper proposes a k-means-type algorithm that is able to provide data clustering and outlier detection simultaneously by incorporating an additional cluster into the objective function and designs an iterative procedure to optimize the objectivefunction of the proposed algorithm and establish the convergence of the Iterative procedure.

TL;DR: A multi-stream convolutional neural network framework to improve the recognition accuracy of gestures by learning the correlation between individual muscles and specific gestures with a “divide-and-conquer” strategy is proposed.

TL;DR: A survey of the recent literature on user profiling and behavioral adaptation in human-robot interaction is presented and introduces a general classification scheme for both the profiling and the behavioral adaptation research topics in terms of physical, cognitive, and social interaction viewpoints.

TL;DR: The experimental results proved that the proposed Bat algorithm (BA) is capable to find the optimal values of the SVM parameters and avoids the local optima problem.

TL;DR: Subjective and objective performance evaluations demonstrate that the proposed method significantly improves both color and visibility of degraded underwater images, and is comparable to and even outperforms several state-of-the-art methods.

TL;DR: An example-driven k-parameter computation that identifies different k values for different test samples in kNN prediction applications, such as classification, regression and missing data imputation is studied.

TL;DR: The results confirm that the features derived using the proposed lexicon outperform those from state-of-the-art lexicons learnt using supervised Latent Dirichlet Allocation (sLDA) and Point-Wise Mutual Information (PMI).

TL;DR: A novel deep learning technique for the recognition of handwritten Bangla isolated compound character is presented and a new benchmark of recognition accuracy on the CMATERdb 3.3.1.3 dataset is reported.

TL;DR: This approach introduces a heart rate measuring strategy using LAB color facial video, which has noteworthy potential for advancing telemedicine, health of a person and numerous applications where information is needed on a real time basis.

TL;DR: A data-level solution has been offered to the concerned problem with novelty in effective elimination of majority instances without losing valuable information by amalgamating aspects of outlier and redundancy detection to the baseline system.

TL;DR: The robust insertion is also optimized with the help of ABC (Artificial Bee colony) in such a way that maximum robustness can be assured corresponding to user specific threshold of imperceptibility.

TL;DR: An automated system for the detection of focal EEG signals using differencing and flexible analytic wavelet transform (FAWT) methods using LS-SVM classifier with ten-fold cross validation strategy is developed.

TL;DR: Using deeply recurrent neural networks to account for temporal dependence in electroencephalograph (EEG)-based workload estimation is shown to considerably improve day-to-day feature stationarity resulting in significantly higher accuracy than classifiers which do not consider the temporal dependence encoded within the EEG time-series signal.

TL;DR: This work proposes extracting sets of spatial and temporal local features from subgroups of joints, which are aggregated by a robust method based on the VLAD algorithm and a pool of clusters and combined by a metric learning method inspired by the LMNN algorithm.

TL;DR: This work formally shows that the GED, restricted to the paths in this family, is equivalent to a quadratic assignment problem, and proposes to compute an approximate solution by adapting two algorithms: Integer Projected Fixed Point method and Graduated Non Convexity and Concavity Procedure.

TL;DR: A taxonomy study for background initialization is proposed, providing the basis for a fair and easy comparison of existing and future methods, on a common dataset of groundtruthed sequences, with a common set of metrics, and based on reproducible results.

TL;DR: The results show that the optical flow information from emotional-face and neutral-face is a useful complement to spatial feature and can effectively improve the performance of facial expression recognition from static images.

TL;DR: The proposed back projection strategy can be extended to other GAN-based image-to-image translation problems and is inspired by the recent success in generating images of generative adversarial networks (GAN).

TL;DR: A new method for face verification across large age gaps and also a dataset containing variations of age in the wild, the Large Age-Gap (LAG) dataset, with images ranging from child/young to adult/old is introduced.

TL;DR: A novel three-stream CNNs architecture able to be used for action feature extraction, and a data augmentation scheme which is very efficient due to its origin at cropping across videos to deal with the inadequacy of training samples during learning.

TL;DR: A convolutional neural network embedding to perform place recognition is introduced and is validated through extensive experimentation that reveals better performance than state-of-the-art methods.

TL;DR: This paper presents a hybrid system where a supervised deep belief network is trained to select generic features, and a kernel-based SVM is trained from the features that learned by the DBN, and substituted linear kernel for nonlinear ones without loss of accuracy.

TL;DR: A novel and effective approach for automated audio classification is presented that is based on the fusion of different sets of features, both visual and acoustic, in an ensemble that produces better classification accuracy than other state-of-the-art approaches.