Feature selection effectively reduces the dimensionality of data. For feature selection, rough set theory offers a systematic theoretical framework based on consistency measures, of which information entropy is one of the most important significance measures of attributes. However, an information-entropy-based significance measure is computationally expensive and requires repeated calculations. Although

The reconstruction of shredded documents consists of coherently arranging fragments of paper (shreds) to recover the original document(s). A great challenge in computational reconstruction is to properly evaluate the compatibility between the shreds. While traditional pixel-based approaches are not robust to real shredding, more sophisticated solutions compromise significantly time performance. The

Many applications in biomedical informatics deal with data in the tensor form. Traditional regression methods which take vectors as covariates may encounter difficulties in handling tensors due to their ultrahigh dimensionality and complex structure. In this paper, we introduce a novel sparse regularized Tucker tensor regression model to exploit the structure of tensor covariates and perform feature

This paper considers the problem of positive unlabeled (PU) learning. In this context, we propose a two-stage GAN-based model. More specifically, the main contribution is to incorporate a biased PU risk within the standard GAN discriminator loss function. In this manner, the discriminator is constrained to steer the generator to converge towards the unlabeled samples distribution while diverging from

Dynamic gesture recognition, which plays an essential role in human-computer interaction, has been widely investigated but not yet fully addressed. The challenge mainly lies in three folders: 1) to model both of the spatial appearance and the temporal evolution simultaneously; 2) to address the interference from the varied and complex background; 3) the requirement of real-time processing. In this

In this paper, we propose a deep relational network which exploits multi-scale information of facial images for kinship verification. Unlike most existing deep learning based facial kinship verification methods which employ convolutional neural networks to extract holistic features, we present a deep model to exploit facial kinship relationship from local regions. For each given pair of face images

Space target recognition is the basic task of space situational awareness and has developed significantly in the last decade. This paper proposes a hybrid convolutional neural network with partial semantic information for space target recognition, which joints the global features and partial semantic information. Firstly, we propose a two-stage target detection network based on the characteristics

Existing domain adaptation approaches focus on taking advantage of easy samples, i.e, target samples which are easier for adaptation. In previous work, tough, or hard, target samples are generally regarded as outliers or just being left to chance. As a result, the adaptation of tough target samples remains as a challenging problem in the community. In this paper, we report three novel ideas for domain

Data heterogeneity such as task heterogeneity, view heterogeneity, and instance heterogeneity often co-exist in many real-world applications including insider threat detection, traffic prediction, brain image analysis, quality control in manufacturing processes, etc. However, most of the existing techniques might not take fully advantage of the rich heterogeneity. To address this problem, we propose

Multitask (multi-target or multi-output) learning (MTL) deals with simultaneous prediction of several outputs. MTL approaches rely on the optimization of a joint score function over the targets. However, defining a joint score in global models is problematic when the target scales are different. To address such problems, single target (i.e. local) learning strategies are commonly employed. Here we

Existing Visible-Thermal Person Re-identification (VT-REID) methods usually adopt two-stream networks for cross-modality images. The two streams are trained to extract features from different modality images respectively. In contrast, we design a Modality Adversarial Neural Network (MANN) to solve VT-REID problem. Our proposed MANN includes a one-stream feature extractor and a modality discriminator

A novel linear multivariate decision tree classifier, Binary Decision Tree based on K-means Splitting (BDTKS), is presented in this paper. The unsupervised K-means clustering is recursively integrated into the binary tree, building a hierarchical classifier. The introduction of the unsupervised K-means clustering provides the powerful generalization ability for the resulting BDTKS model. Then, the

This paper proposes Dirichlet Variational Autoencoder (DirVAE) using a Dirichlet prior. To infer the parameters of DirVAE, we utilize the stochastic gradient method by approximating the inverse cumulative distribution function of the Gamma distribution, which is a component of the Dirichlet distribution. This approximation on a new prior led an investigation on the component collapsing, and DirVAE

Hashing methods for cross-modal retrieval has drawn increasing research interests and has been widely studied in recent years due to the explosive growth of multimedia big data. However, a significant phenomenon which has been ignored is that there is a large gap between the results of cross-modal hashing in most cases. For example, the results of Text-to-Image frequently outperform that of Image-to-Text

Representation learning methods have attracted considerable attention in learning-based face super-resolution in recent years. Conventional methods perform local models learning on low-resolution (LR) manifold and face reconstruction on high-resolution (HR) manifold respectively, leading to unsatisfactory reconstruction performance when the acquired LR face images are severely degraded (e.g., noisy

While Structure from Motion achieves great success in 3D reconstruction, it still meets challenges on large scale scenes. Incremental SfM approaches are robust to outliers, but are limited by low efficiency and easy suffer from drift problem. Though Global SfM methods are more efficient than incremental approaches, they are sensitive to outliers, and would also meet memory limitation and time bottleneck

One-class classification (OCC) is a classical problem in computer vision that can be described as the task of classifying outlier class samples (OC samples) from the OCC model trained on inlier class samples (IC samples) when datasets are highly biased toward one class due to the insufficient sample size of the other class. Currently, the adversarial learning OCC (ALOCC) method has been proven to significantly

In this paper, we aim to address the unsupervised domain adaptation problem where the data in the target domain are much more diverse compared with the data in the source domain. In particular, this problem is formulated as discovering and incorporating latent domains underlying target data of interest for unsupervised domain adaptation. More specifically, the discovery of the latent target domains

In various pattern classification problems, semi-supervised learning methods have shown its effectiveness in utilizing unlabeled data to yield better performance than some supervised and unsupervised learning methods. Semi-supervised discriminant embedding (SDE) is a semi-supervised extension of local discriminant embedding (LDE). However, when dealing with high dimensional data, SDE often suffers

In this paper, we propose a new single shot method for multi-person 3D human pose estimation in complex images. The model jointly learns to locate the human joints in the image, to estimate their 3D coordinates and to group these predictions into full human skeletons. The proposed method deals with a variable number of people and does not need bounding boxes to estimate the 3D poses. It leverages and

The two-stage strategy has been widely used in image classification. However, these methods barely take the classification criteria of the first stage into consideration in the second prediction stage. In this paper, we propose a novel Two-Stage Representation method (TSR), and convert it to a Single-Teacher Single-Student (STSS) problem in our two-stage knowledge transfer framework for image classification

We investigate the intersection between hierarchical and superpixel image segmentation. Two strategies are considered: (i) the classical region merging, that creates a dense hierarchy with a higher number of levels, and (ii) the recursive execution of some superpixel algorithm, which generates a sparse hierarchy with fewer levels. We show that, while dense methods can capture more intermediate or higher-level

Inhomogenous image segmentation has been a research challenge in recent years. To deal with this difficulty, we propose a new local and global active contour model based on Jeffreys divergence. First, unlike the local data fitting energy of the region-scalable fitting model, a new local data fitting energy based on Jeffreys divergence is proposed instead of Euclidean distance, which achieves relatively

Multi-video summarization is an effective tool for users to browse multiple videos. In this paper, multi-video summarization is formulated as a graph analysis problem and a dynamic graph convolutional network is proposed to measure the importance and relevance of each video shot in its own video as well as in the whole video collection. Two strategies are proposed to solve the inherent class imbalance

For performing multi-class classification, deep neural networks almost always employ a One-vs-All (OvA) classification scheme with as many output units as there are classes in a dataset. The problem of this approach is that each output unit requires a complex decision boundary to separate examples from one class from all other examples. In this paper, we propose a novel One-vs-One (OvO) classification

In multi-label classification, each instance is associated with a set of pre-specified labels. One common approach is to use Binary Relevance (BR) paradigm to learn each label by a base classifier separately. Use of k-Nearest Neighbor (kNN) as the base classifier (denoted as BRkNN) is a simple, descriptive and powerful approach. In binary relevance a highly imbalanced view of dataset is used. However

Unconstrained text recognition is an important computer vision task, featuring a wide variety of different sub-tasks, each with its own set of challenges. One of the biggest promises of deep neural networks has been the convergence and automation of feature extractors from input raw signals, allowing for the highest possible performance with minimum required domain knowledge. To this end, we propose

Online multi-object tracking needs to overcome the intrinsic detector deficiencies, e.g., missing detections, false alarms, and inaccurate detection responses, to grow multiple object trajectories without using future information. Various distractions exist during this growing process like background clutters, similar targets, and occlusions, which present a great challenge. We in this work propose

We introduce an approach to divisive hierarchical clustering that is capable of identifying clusters in nonlinear manifolds. This approach uses the isometric mapping (Isomap) to recursively embed (subsets of) the data in one dimension, and then performs a binary partition designed to avoid the splitting of clusters. We provide a theoretical analysis of the conditions under which contiguous and high-density

The semi-supervised video anomaly detection assumes that only normal video clips are available for training. Therefore, the intuitive idea is either to learn a dictionary by sparse coding or to train encoding-decoding neural networks by minimizing the reconstruction errors. For the former, the optimization of sparse coefficients is extremely time-consuming. For the latter, this manner cannot guarantee

Depth image super-resolution (DISR) is an effective solution to improve the quality of depth images captured by real world low-cost cameras. In this paper, we propose a multi-scale symmetric network with the correlation-controlled color guidance block (CCGB) for DISR. The proposed network consists of two multi-scale sub-networks to respectively provide guidance and estimate depth. A symmetric unit

Persistent homology and persistent entropy have recently become useful tools for patter recognition. In this paper, we find requirements under which persistent entropy is stable to small perturbations in the input data and scale invariant. In addition, we describe two new stable summary functions combining persistent entropy and the Betti curve. Finally, we use the previously defined summary functions

Low-rank quaternion tensor completion method, a novel approach to recovery color videos and images, is proposed in this paper. We respectively reconstruct a color image and a color video as a quaternion matrix (second-order tensor) and a third-order quaternion tensor by encoding the red, green, and blue channel pixel values on the three imaginary parts of a quaternion. Different from some traditional

In this study, we attempted to develop a method for accelerating parameter optimization of an object detector ensemble over large image datasets by using simulated annealing. We propose a novel sampling-based evaluation method that considers the minimum portion of the dataset required in each iteration to maintain solution quality. This approach can be considered a noisy evaluation of the energy. The

Variational autoencoders (VAEs), as well as other generative models, have been shown to be efficient and accurate for capturing the latent structure of vast amounts of complex high-dimensional data. However, existing VAEs can still not directly handle data that are heterogenous (mixed continuous and discrete) or incomplete (with missing data at random), which is indeed common in real-world applications

A novel optimization framework for joint unsupervised clustering and kernel learning is derived. Sparse nonnegative matrix factorization of kernel covariance matrices is utilized to categorize data according to their information content. It is demonstrated that a pertinent kernel covariance matrix for clustering can be constructed such that it is block diagonal within arbitrary row and column permutations

This paper proposes a novel image decomposition model for scene depth recovery from low-quality depth measurements and its corresponding high resolution color image. Through our observation, the depth map mainly contains smooth regions separated by additive step discontinuities, and can be simultaneously decomposed into a local smooth surface and an approximately piecewise constant component. Therefore

In this paper, we propose a novel deep learning-based framework for facial landmark detection. This framework takes as input face image returned by a face detector (Faster R-CNN) and generates as output a set of landmarks positions. Prior CNN-based methods often select randomly small local patches to predict an initial guess of landmarks locations. One issue with these local patches is that the adjacent

Objective Clinical and dermoscopy images (multi-modality image pairs) are routinely used sequentially in the assessment of skin lesions. Clinical images characterize a lesion's geometry and color; dermoscopy depicts vascularity, dots and globules from the sub-surface of the lesion. Together these modalities provide labels to characterize a skin lesion. Recently, convolutional neural networks (CNNs)

Handwritten Chinese Character Recognition (HCCR) is a challenging topic in the field of pattern recognition due to large-scale character vocabulary, complex hierarchical structure, various writing styles, and scarce training samples. In this paper, we explored the hierarchical knowledge of Chinese characters and presented a novel zero-shot HCCR method. First, we handled the relations between the characters

The decision-making ability of deep reinforcement learning has been proved successfully in a variety of fields. Here, we use this method for precise character detection by making tight bounding boxes around the Chinese characters in historical documents. An agent is trained to learn the control policy of fine-tuning a bounding box step-by-step through a Markov Decision Process. We introduce a novel

Current orthogonal matching pursuit (OMP) algorithms calculate the correlation between two vectors using the inner product operation and minimize the mean square error, which are both suboptimal when there are non-Gaussian noises or outliers in the observation data. To overcome these problems, a new OMP algorithm is developed based on information theoretic learning (ITL), which is built on the following

For many real-world applications, predicting a price range is more practical and desirable than predicting a concrete value. In this case, price prediction can be regarded as a classification problem. Although deep forest is recognized as the best solution to many classification problems, a crucial issue limits its direct application to price prediction, i.e., it treated all the misclassifications

Multi-task learning is a promising machine learning branch, which aims to improve the generalization of the prediction models by sharing knowledge among tasks. Most of the existing multi-task learning methods rely on predefined task relationships and guide the learning process of models by linear regularization terms. On the one hand, improper setting of task relationships may result in negative knowledge

Traditional manifold learning methods generally include a single projection stage that maps high-dimensional data into lower-dimensional space. However, these methods cannot guarantee that the projection matrix is optimal for classification, which limits their practical application. To address this issue, we propose a two-stage projection matrix optimization model termed self-adaptive manifold discriminant

Skeleton-based action recognition aims to recognize human actions by exploring the inherent characteristics from the given skeleton sequences and has attracted far more attention due to its great important potentials in practical applications. Previous methods have illustrated that learning discriminative spatial and temporal features from the skeleton sequences is a crucial factor to recognize human

Semantic segmentation is an end-to-end task that requires both semantic and spatial accuracy. It is important for deep learning-based segmentation methods to effectively utilize the high-level feature map whose semantic information is abundant and the low-level feature map whose spatial information is accurate. However, existing segmentation networks typically cannot take full advantage of these two

Statistical models have been widely adopted for image set classification owing to their capacity in characterizing the data distribution more flexibly and faithfully. However, these methods typically suffer from the problem that the query image set has weak statistical correlations with the training sets, which leads to larger fluctuations in performance. To address this problem, we propose a semi-supervised

Existing deep trackers can be roughly divided into either matching-based or classification-based methods. The formers are fast but not very robust; while the latter ones introduce more discriminative information but often very slow. In this work, we present a novel real-time robust tracking method to take full use of the benefits from both kinds of networks. First, we propose a matching-classification

Defocus map estimation (DME) is very useful in many computer vision applications and has drawn much attention in recent years. Edge-based DME methods can generate sharp defocus discontinuities but usually suffer from textures of the input image. Region-based methods are free of textures but cannot catch the defocus discontinuities very well. In this paper, we propose a DME method combining edge-based

In this paper, we investigate the problem of Temporal Action Proposal (TAP) generation, which plays a fundamental role in large-scale untrimmed video analysis but remains largely unsolved. Most of the prior works proposed the temporal actions by predicting the temporal boundaries or actionness scores of video units. Nevertheless, context information among surrounding video units has not been adequately

Domain adaptation is a significant and popular issue of solving distribution discrepancy among different domains in computer vision. Generally, previous works proposed are mainly devoted to reducing domain shift between source domain with labeled data and target domain without labels. Adversarial learning in deep networks has already been widely applied to learn disentangled and transferable features

In recent years, fully convolutional neural network (FCN) has broken all records in various vision task. It also achieves great performance in salient object detection. However, most of the state-of-the-art methods have suffered from the challenge of precisely segmenting the entire salient object with uniform region and explicit boundary and effectively suppressing the backgrounds on complex images

I.I.D.5 hypothesis between training and testing data is the basis of numerous image classification methods. Such property can hardly be guaranteed in practice where the Non-IIDness is common, causing instable performances of these models. In literature, however, the Non-I.I.D.6 image classification problem is largely understudied. A key reason is lacking of a well-designed dataset to support related

A challenging issue of clustering in real-word application is to detect clusters with arbitrary shapes and densities in complex data. Many conventional clustering algorithms are capable of detecting non-spherical clusters, but their performance is limited when processing data with complex shapes and multiple density peaks in a cluster without knowing the number of clusters. This paper proposes an adaptive

Biological studies have shown that the interaction between neurons are based on neurotransmitters, which transmit signals between neurons, and that one neuron sends information to another neuron by releasing a number of different neurotransmitters, which play different roles. Motivated by this biological discovery, a novel neural networks model is proposed by extending the dimension of connection weights

Reflection symmetry is a very commonly occurring feature in both natural and man-made objects, which helps in understanding objects better and makes them visually pleasing. Detection of reflection symmetry is a fundamental problem in the field of computer vision and computer graphics which aids in understanding and representing reflective symmetric objects. In this work, we attempt the problem of detecting

In this paper, we propose a spatial-temporal deformable networks approach to investigate both problems of face alignment in static images and face tracking in videos under unconstrained environments. Unlike conventional feature extractions which cannot explicitly exploit augmented spatial geometry for various facial shapes, in our approach, we propose a deformable hourglass networks (DHGN) method,