Aiming at solving the problem of image size limiting in the traditional Random Projection (RP) algorithm, a novel Tighter Random Projection (TRP), which combines the scheme with Minimal Intra-class Variance (TRP-MIV) for hyperspectral remote sensing image classification is proposed. First, a new tighter dimensional boundary for expanding image size with the TRP-MIV matrix selected by multiple sampling

Convolutional neural networks (CNNs) have shown unprecedented success in object representation and detection. Nevertheless, CNNs lack the capability to model context dependencies among objects, which are crucial for salient object detection. As the long short-term memory (LSTM) is advantageous in propagating information, in this paper, we propose two variant LSTM units for the exploration of contextual

In this paper, we propose a speed-up approach for subclass discriminant analysis and formulate a novel efficient multi-view solution to it. The speed-up approach is developed based on graph embedding and spectral regression approaches that involve eigendecomposition of the corresponding Laplacian matrix and regression to its eigenvectors. We show that by exploiting the structure of the between-class

Feature matching plays a very important role in many computer vision and pattern recognition tasks. The spatial neighborhood relationship (representing the topological structures of some key feature points of an image scene) is generally well preserved between two feature points of an image pair. Several mismatch-removing methods that maintain the local neighborhood structures of potential true matches

Many research topics present very high dimensional data. Because of the heavy execution times and large memory requirements, many machine learning methods have difficulty in processing these data. In this paper, we propose a new unsupervised feature selection method considering the pairwise dependence of features (feature dependency-based unsupervised feature selection, or DUFS). To avoid selecting

Nonlinear dimensionality reduction is an active area of research. In this paper, we present a thematically different approach to detect a low-dimensional manifold that lies within a set of bounds derived from a given point cloud. A matrix representing distances on a low-dimensional manifold is low-rank, and our method is based on current low-rank Matrix Completion (MC) techniques for recovering a partially

A novel approach relying on the notion of mixture models is proposed for modeling and clustering directed weighted networks. The developed methodology can be used in a variety of settings including multilayer networks. Computational issues associated with the developed procedure are effectively addressed by the use of MCMC techniques. The utility of the methodology is illustrated on a set of experiments

In this paper, we propose a novel method for projecting data from multiple modalities to a new subspace optimized for one-class classification. The proposed method iteratively transforms the data from the original feature space of each modality to a new common feature space along with finding a joint compact description of data coming from all the modalities. For data in each modality, we define a

Neurophysiological evidence demonstrates that classical receptive field responses in the primary visual cortex can be modulated by the non-classical receptive field. Although models based on the non-classical receptive field have been proposed, which has not employed the two following characteristics: dynamic regulation of the receptive field under external stimuli and depth determination by binocular

Parameters Selection Problem (PSP) is a relevant and complex optimization issue in Support Vector Machine (SVM) and Support Vector Regression (SVR), looking for obtaining an optimal set of hyperparameters. In our case, the optimization problem is addressed to obtain models that minimize the number of support vectors and maximize generalization capacity. However, to obtain accurate and low complexity

Recently, graph neural networks have attracted great attention and achieved prominent performance in various research fields. Most of those algorithms have assumed pairwise relationships of objects of interest. However, in many real applications, the relationships between objects are in higher-order, beyond a pairwise formulation. To efficiently learn deep embeddings on the high-order graph-structured

We propose an improved version of the SMO algorithm for training classification and regression SVMs, based on a Conjugate Descent procedure. This new approach only involves a modest increase on the computational cost of each iteration but, in turn, usually results in a substantial decrease in the number of iterations required to converge to a given precision. Besides, we prove convergence of the iterates

While complex-valued transforms have been widely used in image processing and have their deep connections to biological vision systems, complex-valued convolutional neural networks (CNNs) have not seen their applications in image recovery. This paper aims at investigating the potentials of complex-valued CNNs for image denoising. A CNN is developed for image denoising with its key mathematical operations

Despite the promising preliminary results, none of existing deep learning based cross-view classification methods simultaneously takes into account both view consistency learning and class-specificity distribution of the extracted features, resulting in unstable classification performance. Moreover, most existing cross-view classification methods are sensitive to scale due to the scale issue of view

In this paper, a new discriminative dictionary learning algorithm is introduced. An entropy based criterion is embedded into the objective function to enforce a proper structure for the dictionary items when decomposing signals of different classes. The proposed criterion influences the dictionary items to participate in the decomposition of a smaller number of classes as possible. Unlike the other

Deep metric learning (DML) has been designed to maximize the inter-class variance that is the distance between embedding features belonging to different classes. Since conventional DML techniques do not consider the statistical characteristics of the embedding space, or they calculate similarity using only a given feature, they make it difficult to adaptively reflect the characteristics of the feature

Lane detection is an application of environmental perception, which aims to detect lane areas or lane lines by camera or lidar. In recent years, gratifying progress has been made in detection accuracy. To the best of our knowledge, this paper is the first attempt to make a comprehensive review of vision-based lane detection methods. First, we introduce the background of lane detection, including traditional

Trajectory prediction is a crucial element of many automated tasks, such as autonomous navigation or video surveillance. To automatically predict the motion of an agent (e.g., pedestrian or car), the model needs to efficiently represent human motion and “understand” the external stimuli that may influence human behavior. In this work we propose a methodology to model the motion of agents in a video

In this paper, we focus on event discovery by utilizing data distributed in multiple media domains, such as news media and social media. To this end, we propose an in-domain and cross-domain Laplacian regularization (ICLR) model, which can learn effective data representation for both textual news reports contributed by journalists in news media domain, and image posts shared by amateur users in social

Unsupervised domain adaptation is an effective approach to solve the problem of dataset bias. However, most existing unsupervised domain adaptation methods assume that the geometry structures of data distributions are similar in the source and target domains. This assumption is invalid in many practical applications, because the training and test datasets usually differ in the variability modes and/or

For one given scene, multimodal data are acquired from multiple sensors. They share some similarities across the sensor types (redundant part of the information, also called coupling part) and they also provide modality-specific information (dissimilarities across the sensors, also called decoupling part). Additional critical knowledge about the scene can hence be extracted, which is not extractable

One primary challenge of face recognition is that the performance is seriously affected by varying illumination. Multi-spectral imaging can capture face images in the visible spectrum and beyond, which is deemed to be an effective technology in response to this challenge. For current multi-spectral imaging-based face recognition methods, how to fully explore the discriminant and correlation features

Generative Adversarial Networks (GANs) have been widely used to generate realistic-looking instances. However, training robust GAN is a non-trivial task due to the problem of mode collapse. Although many GAN variants are proposed to overcome this problem, they have limitations. Those existing studies either generate identical instances or result in negative gradients during training. In this paper

Atrous Spatial Pyramid Pooling (ASPP) is a module that can collect semantic information distributed in different scopes. However, because of the limited number of sampling ranges of ASPP, much valuable global features and contextual information cannot be sufficiently sampled, which degrades the representation ability of the segmentation network. Besides, due to the sparse distribution of the effective

The problem of semi-supervised video object segmentation has been popularly tackled by fine-tuning a general-purpose segmentation deep network on the annotated frame using hundreds of iterations of gradient descent. The time-consuming fine-tuning process, however, makes these methods difficult to use in practical applications. We propose a novel architecture called Annotation Guided U-net (AGUnet)

The key to the success of an Error-Correcting Output Code (ECOC) algorithm is the effective codematrix, which represents a set of class reassignment schemes for decomposing a multiclass problem into a set of binary class problems. This paper proposes a new method, which uses Ternary digit Operators based Genetic Programming (GP) to generate effective ECOC codematrix (TOGP-ECOC for short). In our GP

Sequence labeling is a fundamental task in natural language processing and has been widely studied. Recently, RNN-based sequence labeling models have increasingly gained attentions. Despite superior performance achieved by learning the long short-term (i.e., successive) dependencies, the way of sequentially processing inputs might limit the ability to capture the non-continuous relations over tokens

Emotion plays a vital role in human daily life, and EEG signals are widely used in emotion recognition. Due to individual variability, training a generic emotion recognition model across different subjects is difficult. The conventional method involves the collection of a large amount of calibration data to build subject-specific models. Recently, developing an effective brain-computer interface with

Domain adaptation aims to learn an adaptive classifier for target data using the labelled source data from a different distribution. Most proposed works construct cross-domain classifier by exploring one-sided property of the input data, i.e., either geometric or statistical property. Therefore they may ignore the complementarity between the two properties. Moreover, many previous methods implement

Knowledge distillation is an effective way to transfer the knowledge from a pre-trained teacher model to a student model. Co-distillation, as an online variant of distillation, further accelerates the training process and paves a new way to explore the “dark knowledge” by training n models in parallel. In this paper, we explore the “divergent examples”, which can make the classifiers have different

In this work, we address the task of scene text retrieval: given a text query, the system returns all images containing the queried text. The proposed model uses a single shot CNN architecture that predicts bounding boxes and builds a compact representation of spotted words. In this way, this problem can be modeled as a nearest neighbor search of the textual representation of a query over the outputs

Feature Selection (FS) plays an important role in learning and classification tasks. Its objective is to select the relevant and non-redundant features. Considering the huge number of features in real-world applications, FS methods using batch learning technique cannot resolve big data problems especially when data arrive sequentially. In this paper, we proposed an online feature selection system which

Existing high-performance object detection methods greatly benefit from the powerful representation ability of deep convolutional neural networks (CNNs). Recent researches show that integration of high-order statistics remarkably improves the representation ability of deep CNNs. However, high-order statistics for object detection lie in two challenges. Firstly, previous methods insert high-order statistics

Graphs have become increasingly popular in modeling structures and interactions in a wide variety of problems during the last decade. Graph-based clustering and semi-supervised classification techniques have shown impressive performance. This paper proposes a graph learning framework to preserve both the local and global structure of data. Specifically, our method uses the self-expressiveness of samples

The K-means clustering algorithm is well-known for its easy computational approach. In this algorithm, essential cluster-level information is captured by the K cluster centroids. However, how many such centroids can reveal the structure of the underlying data depends upon the choice of K. In this paper, we propose a clustering algorithm in which the number of cluster K can be learned as well as it

The studies of hidden complex structures in data have popularized the use of graph-based learning methods in semi-supervised and unsupervised learning tasks. Kernelized graph-based methods are proven to perform better, but these methods suffer from the issue of appropriate kernel selection. Instead of using multiple views, these methods generally use a single view. But multi-view methods need a proper

Deep spatio-temporal neural networks have shown promising performance for video super-resolution (VSR) in recent years. However, most of them heavily rely on accuracy motion estimations. In this paper, we propose a novel spatio-temporal matching network (STMN) for video super-resolution, which works on the wavelet domain to reduce dependence on motion estimations. Specifically, our STMN consists of

Due to a variety of motions across different frames, it is highly challenging to learn an effective spatiotemporal representation for accurate video saliency prediction (VSP). To address this issue, we develop an effective spatiotemporal feature alignment network tailored to VSP, mainly including two key sub-networks: a multi-scale deformable convolutional alignment network (MDAN) and a bidirectional

Circle detection is a critical issue in pattern recognition and image analysis. Conventional geometry-based methods such as tangent or symmetry are sensitive to noise or occlusion. Area computation is more robust against noise, because it avoids differential calculations. Inspired by this characteristic, we present a novel method for fast circle detection using inscribed triangles. The proposed algorithm

Due to the high cost of manual annotation, learning directly from the web has attracted broad attention. One issue that limits the performance of current webly supervised models is the problem of visual polysemy. In this work, we present a novel framework that resolves visual polysemy by dynamically matching candidate text queries with retrieved images. Specifically, our proposed framework includes

We analyze the drawbacks of DBSCAN and its variants, and find the grid technique, which is used in Fast-DBSCAN and ρ-approximate DBSCAN, is almost useless in high dimensional data space. Because it usually yields considerable redundant distance computations. In order to tame these problems, two techniques are proposed: one is to use ϵ2-norm ball to identify Inner Core Blocks within which all points

Dewarping is a necessary preprocessing step to recognize text from a distorted camera captured document image. According to recent literature, deep learning-based approaches perform with higher accuracy in similar domains. The deep learning-based neural networks are not yet fully explored in the domain of dewarping. To fill this gap, we propose a dewarping approach based on the convolutional neural

Most existing crowd counting methods require object location-level annotation which is labor-intensive and time-consuming to obtain. In contrast, weaker annotations that only label the total count of objects can be easy to obtain in many practical scenarios. This paper focuses on the problem of weakly-supervised crowd counting which learns a model from a small amount of location-level annotations (fully-supervised)

Due to the extensive practical value of time series prediction, many excellent algorithms have been proposed. Most of these methods are developed assuming that massive labeled training data are available. However, this assumption might be invalid in some actual situations. To address this limitation, a transfer learning framework with deep architectures is proposed. Since convolutional neural network

Visible-near infrared (VIS-NIR) face matching is a challenging issue in heterogeneous face recognition due to the large spectrum domain discrepancy as well as the over-fitting on insufficient pairwise VIS and NIR images during training. This paper proposes a coupled adversarial learning (CAL) approach for the VIS-NIR face matching by performing adversarial learning on both image and feature levels

Existing monocular depth estimation methods are unsatisfactory due to the inaccurate inference of depth details and the loss of spatial information. In this paper, we present a novel detail-preserving network (DPNet), i.e., a dual-branch network architecture that fully addresses the above problems and facilitates the depth map inference. Specifically, in contextual branch (CB), we propose an effective

The COVID-19 outbreak continues to threaten the health and life of people worldwide. It is an immediate priority to develop and test a computer-aided detection (CAD) scheme based on deep learning (DL) to automatically localize and differentiate COVID-19 from community-acquired pneumonia (CAP) on chest X-rays. Therefore, this study aims to develop and test an efficient and accurate deep learning scheme

Deep learning has received increasing attention in the last decade. Its amazing success, is partly attributed to the evolution of normalization and activation techniques. However, less works have devoted to explore both modules together. This work, therefore, aims at pushing for a deeper understanding on the effect of normalization and activation together analytically. We design a generic method which

In this paper, we propose a dual subspace discriminative projection learning (DSDPL) framework for multi-category image classification. Our approach reflects the notion that images are composed of class-shared information, class-specific information, and sparse noise. Unlike traditional subspace learning methods, DSDPL serves to decompose original high dimensional data, via learned projection matrices

In this paper, we aim to tackle the one-shot person re-identification problem where only one image is labelled for each person, while other images are unlabelled. This task is challenging due to the lack of sufficient labelled training data. To tackle this problem, we propose to iteratively guess pseudo labels for the unlabelled image samples, which are later used to update the re-identification model

Convolutional neural networks (CNNs) have been successfully applied in many computer vision applications [1], especially in image classification tasks, where most of the structures have been designed manually. With the aid of skip connection and dense connection, the depths of the models are becoming “deeper” and the filters of layers are getting “wider” in order to tackle the challenge of large-scale

The capability of multi column convolutional networks in identifying local invariant features helps improve its performance on image classification tasks to a large extent. Suppression of non maximal activations in a convolutional network, however, can lead to loss of valuable information, as scalar activations typically only ,encode the presence (or absence) of a feature in an input image, providing

Kernel ridge regression (KRR) is an efficient method for regression task. However, KRR has a deficiency in finding appropriate type of kernel functions and their parameters. To overcome this shortcoming, a novel kernel ensemble framework is developed. In this ensemble framework, each kernel regressor is associated with a weight that can be adaptively determined according to its contribution to the

In computer vision, the research community has been looking to how to benefit from weakly supervised learning that utilizes easily obtained image-level labels to train neural network models. The existing deep convolutional neural networks for weakly supervised learning, however, generally do not fully exploit the label dependencies in an image. To make full use of this information, in this paper, we

The random k-labelsets ensemble (RAkEL) is a multi-label learning strategy that integrates many single-label learning models. Each single-label model is constructed using a label powerset (LP) technique based on a randomly generated size-k label subset. Although RAkEL can improve the generalization capability and reduce the complexity of the original LP method, the quality of the randomly generated

When the labelling information is not deterministic, traditional supervised learning algorithms cannot be applied. In this case, stochastic supervision models provide a valuable alternative to classification. However, these models are restricted in several aspects, which critically limits their applicability. In this paper, we provide four generalisations of stochastic supervision models, extending

Blind image deblurring is a very challenging inverse problem due to the severe ill-posedness caused by the unknown kernel and the latent clear image. To tackle this problem, appropriate smoothing regularizations and image priors are usually employed and incorporated into the associated variational models to alleviate the inherent ill-posedness. In this paper, we first propose a strongly imposed zero