We present a novel boosting cascade based face detection framework using SURF features. The is derived from the well-known Viola-Jones (VJ) but distinguished by two key contributions. First, proposed deals with only several hundreds of multidimensional local patches instead thousands single dimensional haar features in VJ framework. Second, it takes AUC as criterion for convergence test each stage rather than conflicting criteria (false-positive-rate and detection-rate) These modifications...

In the past few decades, to reduce risk of X-ray in computed tomography (CT), low-dose CT image denoising has attracted extensive attention from researchers, which become an important research issue field medical images. recent years, with rapid development deep learning technology, many algorithms have emerged apply convolutional neural networks this task, achieving promising results. However, there are still some problems such as low efficiency, over-smoothed result, etc. paper, we propose...

Multi-Label Learning (MLL) aims to learn from the training data where each example is represented by a single instance while associated with set of candidate labels. Most existing MLL methods are typically designed handle problem missing However, in many real-world scenarios, labeling information for multi-label always redundant , which can not be solved classical methods, thus novel Partial Multi-label (PML) framework proposed cope such problem, i.e. removing noisy labels sets. In this...

Partial Label Learning (PLL) aims to learn from the data where each training example is associated with a set of candidate labels, among which only one correct. The key deal such problem disambiguate label sets and obtain correct assignments between instances their labels. In this paper, we interpret as instance-to-label matchings, reformulate task PLL matching selection problem. To model problem, propose novel Graph Matching based (GM-PLL) framework, (GM) scheme incorporated owing its...

Matching-based algorithms have been commonly used in planar object tracking. They often model a as set of keypoints, and then find correspondences between keypoint sets via descriptor matching. In previous work, unary constraints on appearances or locations are usually to guide the However, these approaches rarely utilize structure information object, thus suffering from various perturbation factors. this paper, we proposed graph-based tracker, named Gracker, which is able fully explore...

Downsampling is a crucial task for processing large scale and/or dense point clouds with limited resources. Owing to the development of deep learning, approaches task-oriented cloud downsampling have significant performance gains in preserving geometric information. However, most downsamling methods are by disordered and unstructured data, making it difficult continually improve performance. To address this issue, we propose light-weight Transformer network (LighTN) as an end-to-end...

Graph matching aims at establishing correspondences between graph elements, and is widely used in many computer vision tasks. Among recently proposed algorithms, those utilizing the path following strategy have attracted special research attentions due to their exhibition of state-of-the-art performances. However, paths computed these algorithms often contain singular points, which could hurt performance if not dealt properly. To deal with this issue, we propose a novel strategy, named...

Learning-based approaches to graph matching have been developed and explored for more than a decade, grown rapidly in scope popularity recent years. However, previous learning-based algorithms, with or without deep learning strategy, mainly focus on the of node and/or edge affinities generation, pay less attention combinatorial solver. In this paper we propose fully trainable framework matching, which solving optimization are not explicitly separated as many arts. We firstly convert problem...

Human facial age estimation has attracted much attention due to its potential applications in forensics, security, and biometrics. In contrast existing approaches that cast as either a multiclass classification or regression problem, this work, we propose novel approach combines the strength of cost-sensitive label ranking methods with power low-rank matrix recovery theories. Instead having make binary decision for each label, our ranks labels descending order terms their predicted relevance...

With the increasing prevalence of intelligent traffic control and monitoring, research on vehicle re-identification (Re-ID) draws substantial attention in recent years. Different from other cross-view searching tasks such as person Re-ID, Re-ID problem is more challenging unpredictable viewpoint variations can greatly affect appearance vehicles. Existing studies mainly focus extracting global features based visual to represent identity target vehicle, while impact variation rarely...

This paper proposes a general framework called Markov stationary features (MSF) to extend histogram based features. The MSF characterizes the spatial co-occurrence of patterns by chain models, and finally yields compact feature representation through analysis. Therefore, goes one step beyond histograms since it now involves structure information both within bins between bins. Moreover, still keeps simplicity, compactness, efficiency, robustness. We demonstrate how is used like color...

Abstract In knowledge graph embedding, multidimensional representations of entities and relations are learned in vector space. Although distance-based embedding methods have shown promise link prediction, they neglect context information among the triplet components, i.e., head_entity, relation, tail_entity, limiting their ability to describe multivariate relation patterns mapping properties. Such denotes entity structural association inside same implies correlation between that not directly...

Visible-Infrared person Re-IDentification (VI-ReID) is a challenging cross-modality image retrieval task that aims to match pedestrians' images across visible and infrared cameras. To solve the modality gap, existing mainstream methods adopt learning paradigm converting into an classification with cross-entropy loss auxiliary metric losses. These losses follow strategy of adjusting distribution extracted embeddings reduce intra-class distance increase inter-class distance. However, such...

The advent of the information and intelligence era has led to explosive growth data. author proposes a hybrid data model based on differential privacy. main content this is study privacy, processing through noise mechanism, using calculation tuple attribute differences addition, finally constructing mixed privacy experiments. experimental results indicate that: as value k increases, clustering tend be optimal, verifying that original can reduce addition. However, ICMD-DP anonymizes dataset,...