Regular papers
Conditional convolution neural network enhanced random forest for facial expression recognition
- A conditional convolution neural network enhanced random forests for facial expression recognition.
In real-world applications, factors such as head pose variation, occlusion, and poor image quality make facial expression recognition (FER) an open challenge. In this paper, a novel conditional convolutional neural network enhanced ...
Face recognition from multiple stylistic sketches: Scenarios, datasets, and evaluation
- A fundamental study focusing on face recognition from multiple stylistic sketches is presented, and three specific scenarios with corresponding datasets are ...
Due to the great texture discrepancies between photos and sketches, matching face sketches against mug shot photos is a challenging yet important topic in face recognition community, with potential applications in law enforcement and ...
Subject-Specific prior shape knowledge in feature-oriented probability maps for fully automatized liver segmentation in MR volume data
- Fully automatized liver segmentation approach for native MR volume data
- Subject-...
Liver segmentation and volumetry in native MR-volume data is an important topic in epidemiological research. Manual liver segmentation is extremely time-consuming and often infeasible requiring automatized methods. Automatic liver ...
Graphical abstractDisplay Omitted
Angular Radon spectrum for rotation estimation
- Formal definition and derivation of Angular Randon Spectrum (ARS).
- Estimation ...
This paper presents a robust method for rotation estimation of planar point sets using the Angular Radon Spectrum (ARS). Given a Gaussian Mixture Model (GMM) representing the point distribution, the ARS is a continuous function derived ...
Affinity learning via a diffusion process for subspace clustering
- An effective affinity learning method is proposed.
- This approach utilizes the ...
Subspace clustering refers to the problem of finding low-dimensional subspaces (clusters) for high-dimensional data. Current state-of-the-art subspace clustering methods are usually based on spectral clustering, where an affinity ...
A safe sample screening rule for Laplacian twin parametric-margin support vector machine
- A safe sample screening rule (SSSR-LTPSVM) for the LTPSVM is presented based on variational inequality.
Laplacian support vector machine (SVM) for semi-supervised classification has attracted much attention in recent years. As an extension to improve the computational speed, Laplacian twin parametric-margin SVM (LTPSVM) has shown ...
Morphological classifiers
- Classification framework proposal that uses mathematical morphology.
- ...
This work proposes a new type of classifier called Morphological Classifier (MC). MCs aggregate concepts from mathematical morphology and supervised learning. The outcomes of this aggregation are classifiers that may preserve shape ...
A novel formulation of orthogonal polynomial kernel functions for SVM classifiers: The Gegenbauer family
- A novel formulation of orthogonal kernels for SVM classifiers is presented.
- The ...
Orthogonal polynomial kernels have been recently introduced to enhance support vector machine classifiers by reducing their number of support vectors. Previous works have studied these kernels as isolated cases and discussed only ...
Online multi-label streaming feature selection based on neighborhood rough set
- A new neighborhood relation is proposed to effectively solve the problem of granularity selection in neighborhood rough set.
Multi-label feature selection has grabbed intensive attention in many big data applications. However, traditional multi-label feature selection methods generally ignore a real-world scenario, i.e., the features constantly flow into the ...
Distinctiveness, complexity, and repeatability of online signature templates
- The metric to measure distinctiveness, complexity, and repeatability of an online signature template are proposed.
This paper proposes three measures to quantify the characteristics of online signature templates in terms of distinctiveness, complexity and repeatability. A distinctiveness measure of a signature template is computed from a set of ...
F-NSP+: A fast negative sequential patterns mining method with self-adaptive data storage
- We propose a novel and efficient data structure, bitmap, to obtain the support of Negative sequential candidates (NSC).
Mining negative sequential patterns (NSP) is an important tool for nonoccurring behavior analysis, and it is much more challenging than mining positive sequential patterns (PSPs) due to the high computational complexity and huge search ...
A novel framework for background subtraction and foreground detection
- A new data structure, Mino Vector (MV), is designed to serve the whole framework.
In this paper, we propose a novel image background subtraction framework based on KDE. Firstly a new data structure called Mino Vector (MV) is designed for each pixel; we define dynamic nature (DN) for pixels of a scene and rank them ...
Weakly-supervised object detection via mining pseudo ground truth bounding-boxes
- A novel W2F framework for weakly-supervised object detection is proposed.
- The ...
Recently, weakly-supervised object detection has attracted much attention, since it does not require expensive bounding-box annotations while training the network. Although significant progress has also been made, there is still a ...
Circular mesh-based shape and margin descriptor for object detection
- A dynamic circular mesh-based shape and margin descriptor (CMSMD) is proposed.
- ...
In many engineering and medical imaging fields, shape, and margin descriptors play an important role in challenging pattern recognition and classification problems. This paper presents a circular mesh-based shape and margin descriptor (...
Graphical abstractDisplay Omitted
Parametric and nonparametric context models: A unified approach to scene parsing
- A new nonparametric scene parsing approach is proposed.
- Parametric and ...
In this paper a new nonparametric scene parsing approach is proposed which has three steps: image retrieval, label transferring and label gathering. In our approach, to incorporate the contextual knowledge in scene parsing, we propose ...
Learning spatial relations and shapes for structural object description and scene recognition
- We propose a graph-based multilevel descriptor for object recognition.
- This ...
Being able to describe the content of an image, adapted to a particular application, is essential in various domains related to image analysis and pattern recognition. In this context, taking into account the spatial organization of ...
MoE-SPNet: A mixture-of-experts scene parsing network
Scene parsing is an indispensable component in understanding the semantics within a scene. Traditional methods rely on handcrafted local features and probabilistic graphical models to incorporate local and global cues. Recently, ...
Binary Partition Tree construction from multiple features for image segmentation
- We propose a multi-feature BPT construction framework for image segmentation purpose.
In the context of image analysis, the Binary Partition Tree (BPT) is a classical data structure for the hierarchical modelling of images at different scales. BPTs belong both to the families of graph-based models and morphological ...
A fast robust geometric fitting method for parabolic curves
- Ezequiel López-Rubio,
- Karl Thurnhofer-Hemsi,
- Elidia Beatriz Blázquez-Parra,
- Óscar David de Cózar-Macías,
- M. Carmen Ladrón-de-Guevara-Muñoz
- A new method to fit parabolic curves is presented.
- The method is fast and ...
Fitting discrete data obtained by image acquisition devices to a curve is a common task in many fields of science and engineering. In particular, the parabola is some of the most employed shape features in electrical engineering and ...
Using fuzzy least squares support vector machine with metric learning for object tracking
- A new FLS-SVM-ML algorithm is proposed.
- A two-stage iterative process is used ...
Some researchers have introduced the fuzzy learning into tracking and the kernelized fuzzy least squares support vector machine (FLS-SVM) has achieved great success in building the appearance model. However, the kernel used in FLS-SVM ...
Multi-view label embedding
- This paper presents a novel multi-view label embedding algorithm via latent space learning.
Multi-label classification has been successfully applied to image annotation, information retrieval, text categorization, etc. When the number of classes increases significantly, the traditional multi-label learning models will become ...
Model-based learning for point pattern data
This article proposes a framework for model-based point pattern learning using point process theory. Likelihood functions for point pattern data derived from point process theory enable principled yet conceptually transparent ...
Robust one-class support vector machine with rescaled hinge loss function
- A novel robust one-class support vector machine (OCSVM) based on the rescaled hinge loss function is proposed;
In this paper, a novel robust one-class support vector machine (OCSVM) based on the rescaled hinge loss function is proposed to enhance the robustness of the conventional OCSVM against outliers. The optimization problem of the proposed ...
Learning visual and textual representations for multimodal matching and classification
- A unified network for image-text matching and classification.
- Seamlessly ...
Multimodal learning has been an important and challenging problem for decades, which aims to bridge the modality gap between heterogeneous representations, such as vision and language. Unlike many current approaches which only focus on ...
Detection and classification of cancer in whole slide breast histopathology images using deep convolutional networks
- Commonly studied scenario considers only binary cancer vs. no cancer classification.
Generalizability of algorithms for binary cancer vs. no cancer classification is unknown for clinically more significant multi-class scenarios where intermediate categories have different risk factors and treatment strategies. We ...
Correntropy-based robust multilayer extreme learning machines
In extreme learning machines (ELM), the hidden node parameters are randomly generated and the output weights can be analytically computed. To overcome the bad feature extraction ability of the shallow architecture of ELM, the ...
Wild patterns: Ten years after the rise of adversarial machine learning
- We provide a detailed review of the evolution of adversarial machine learning over the last ten years.
Learning-based pattern classifiers, including deep networks, have shown impressive performance in several application domains, ranging from computer vision to cybersecurity. However, it has also been shown that adversarial input ...