Scene text detection plays a significant role in various applications, such as object recognition, document management, and visual navigation. The instance segmentation based method has been mostly used in existing research due to its advantages in dealing with multi-oriented texts. However, a large number of non-text pixels exist in the labels during the model training, leading to text mis-segmentation. In this paper, we propose a novel multi-oriented scene text detection framework, which includes two main modules:character instance segmentation (one instance corresponds to one character), and character flow construction (one character flow corresponds to one word). We use feature pyramid network (FPN) to predict character and non-character instances with arbitrary directions. A joint network of FPN and bidirectional long short-term memory (BLSTM) is developed to explore the context information among isolated characters, which are finally grouped into character flows. Extensive experiments are conducted on ICDAR2013, ICDAR2015, MSRA-TD500 and MLT datasets to demonstrate the effectiveness of our approach. The F-measures are 92.62%, 88.02%, 83.69% and 77.81%, respectively.

Existing physical cloth simulators suffer from expensive computation and difficulties in tuning mechanical parameters to get desired wrinkling behaviors. Data-driven methods provide an alternative solution. They typically synthesize cloth animation at a much lower computational cost, and also create wrinkling effects that are similar to the training data. In this paper we propose a deep learning based method for synthesizing cloth animation with high resolution meshes. To do this we first create a dataset for training:a pair of low and high resolution meshes are simulated and their motions are synchronized. As a result the two meshes exhibit similar large-scale deformation but different small wrinkles. Each simulated mesh pair is then converted into a pair of low- and high-resolution "images" (a 2D array of samples), with each image pixel being interpreted as any of three descriptors:the displacement, the normal and the velocity. With these image pairs, we design a multi-feature super-resolution (MFSR) network that jointly trains an upsampling synthesizer for the three descriptors. The MFSR architecture consists of shared and task-specific layers to learn multi-level features when super-resolving three descriptors simultaneously. Frame-to-frame consistency is well maintained thanks to the proposed kinematics-based loss function. Our method achieves realistic results at high frame rates:12-14 times faster than traditional physical simulation. We demonstrate the performance of our method with various experimental scenes, including a dressed character with sophisticated collisions.

Indoor visual localization, i.e., 6 Degree-of-Freedom camera pose estimation for a query image with respect to a known scene, is gaining increased attention driven by rapid progress of applications such as robotics and augmented reality. However, drastic visual discrepancies between an onsite query image and prerecorded indoor images cast a significant challenge for visual localization. In this paper, based on the key observation of the constant existence of planar surfaces such as floors or walls in indoor scenes, we propose a novel system incorporating geometric information to address issues using only pixelated images. Through the system implementation, we contribute a hierarchical structure consisting of pre-scanned images and point cloud, as well as a distilled representation of the planar-element layout extracted from the original dataset. A view synthesis procedure is designed to generate synthetic images as complementary to that of a sparsely sampled dataset. Moreover, a global image descriptor based on the image statistic modality, called block mean, variance, and color (BMVC), was employed to speed up the candidate pose identification incorporated with a traditional convolutional neural network (CNN) descriptor. Experimental results on a popular benchmark demonstrate that the proposed method outperforms the state-of-the-art approaches in terms of visual localization validity and accuracy.

Face analysis tasks, e.g., estimating gender or age from a face image, have been attracting increasing interest in recent years. However, most existing studies focus mainly on analyzing an adult's face and ignore an interesting question:is it easy to estimate gender and age from a baby's face? In this paper, we explore this interesting problem. We first collect a new face image dataset for our research, named BabyFace, which contains 15 528 images from 5 872 babies younger than two years old. Besides gender, each face image is annotated with age in months from 0 to 24. In addition, we propose new age estimation and gender recognition methods. In particular, based on SSR-Net backbone, we introduce the attention mechanism module to solve the age estimation problem on the BabyFace dataset, named SSR-SE. In the part of gender recognition, inspired by the age estimation method, we also use a two-stream structure, named Two-Steam SE-block with Augment (TSSEAug). We extensively evaluate the performance of the proposed methods against the state-of-the-art methods on BabyFace. Our age estimation model achieves very appealing performance with an estimation error of less than two months. The proposed gender recognition method obtains the best accuracy among all compared methods. To the best of our knowledge, we are the first to study age estimation and gender recognition from a baby's face image, which is complementary to existing adult gender and age estimation methods and can shed some light on exploring baby face analysis.

3D shape editing is widely used in a range of applications such as movie production, computer games and computer aided design. It is also a popular research topic in computer graphics and computer vision. In past decades, researchers have developed a series of editing methods to make the editing process faster, more robust, and more reliable. Traditionally, the deformed shape is determined by the optimal transformation and weights for an energy formulation. With increasing availability of 3D shapes on the Internet, data-driven methods were proposed to improve the editing results. More recently as the deep neural networks became popular, many deep learning based editing methods have been developed in this field, which are naturally data-driven. We mainly survey recent research studies from the geometric viewpoint to those emerging neural deformation techniques and categorize them into organic shape editing methods and man-made model editing methods. Both traditional methods and recent neural network based methods are reviewed.

The 3D object tracking from a monocular RGB image is a challenging task. Although popular color and edgebased methods have been well studied, they are only applicable to certain cases and new solutions to the challenges in real environment must be developed. In this paper, we propose a robust 3D object tracking method with adaptively weighted local bundles called AWLB tracker to handle more complicated cases. Each bundle represents a local region containing a set of local features. To alleviate the negative effect of the features in low-confidence regions, the bundles are adaptively weighted using a spatially-variant weighting function based on the confidence values of the involved energy terms. Therefore, in each frame, the weights of the energy items in each bundle are adapted to different situations and different regions of the same frame. Experiments show that the proposed method can improve the overall accuracy in challenging cases. We then verify the effectiveness of the proposed confidence-based adaptive weighting method using ablation studies and show that the proposed method overperforms the existing single-feature methods and multi-feature methods without adaptive weighting.

In multi-view stereo, unreliable matching in low-textured regions has a negative impact on the completeness of reconstructed models. Since the photometric consistency of low-textured regions is not discriminative under a local window, non-local information provided by the Markov Random Field (MRF) model can alleviate the matching ambiguity but is limited in continuous space with high computational complexity. Owing to its sampling and propagation strategy, PatchMatch multi-view stereo methods have advantages in terms of optimizing the continuous labeling problem. In this paper, we propose a novel method to address this problem, namely the Coarse-Hypotheses Guided Non-Local PAtchMatch Multi-View Stereo (CNLPA-MVS), which takes the advantages of both MRF-based non-local methods and PatchMatch multi-view stereo and compensates for their defects mutually. First, we combine dynamic programing (DP) and sequential propagation along scanlines in parallel to perform CNLPA-MVS, thereby obtaining the optimal depth and normal hypotheses. Second, we introduce coarse inference within a universal window provided by winner-takes-all to eliminate the stripe artifacts caused by DP and improve completeness. Third, we add a local consistency strategy based on the hypotheses of similar color pixels sharing approximate values into CNLPA-MVS for further improving completeness. CNLPA-MVS was validated on public benchmarks and achieved state-of-the-art performance with high completeness.

Partial label learning is a weakly supervised learning framework in which each instance is associated with multiple candidate labels, among which only one is the ground-truth label. This paper proposes a unified formulation that employs proper label constraints for training models while simultaneously performing pseudo-labeling. Unlike existing partial label learning approaches that only leverage similarities in the feature space without utilizing label constraints, our pseudo-labeling process leverages similarities and differences in the feature space using the same candidate label constraints and then disambiguates noise labels. Extensive experiments on artificial and real-world partial label datasets show that our approach significantly outperforms state-of-the-art counterparts on classification prediction.

Unsupervised domain adaptation (UDA) has achieved great success in handling cross-domain machine learning applications. It typically benefits the model training of unlabeled target domain by leveraging knowledge from labeled source domain. For this purpose, the minimization of the marginal distribution divergence and conditional distribution divergence between the source and the target domain is widely adopted in existing work. Nevertheless, for the sake of privacy preservation, the source domain is usually not provided with training data but trained predictor (e.g., classifier). This incurs the above studies infeasible because the marginal and conditional distributions of the source domain are incalculable. To this end, this article proposes a source-free UDA which jointly models domain adaptation and sample transport learning, namely Sample Transport Domain Adaptation (STDA). Specifically, STDA constructs the pseudo source domain according to the aggregated decision boundaries of multiple source classifiers made on the target domain. Then, it refines the pseudo source domain by augmenting it through transporting those target samples with high confidence, and consequently generates labels for the target domain. We train the STDA model by performing domain adaptation with sample transport between the above steps in alternating manner, and eventually achieve knowledge adaptation to the target domain and attain confident labels for it. Finally, evaluation results have validated effectiveness and superiority of the proposed method.

Electrocardiogram (ECG) biometric recognition has emerged as a hot research topic in the past decade. Although some promising results have been reported, especially using sparse representation learning (SRL) and deep neural network, robust identification for small-scale data is still a challenge. To address this issue, we integrate SRL into a deep cascade model, and propose a multi-scale deep cascade bi-forest (MDCBF) model for ECG biometric recognition. We design the bi-forest based feature generator by fusing L1-norm sparsity and L2-norm collaborative representation to efficiently deal with noise. Then we propose a deep cascade framework, which includes multi-scale signal coding and deep cascade coding. In the former, we design an adaptive weighted pooling operation, which can fully explore the discriminative information of segments with low noise. In deep cascade coding, we propose level-wise class coding without backpropagation to mine more discriminative features. Extensive experiments are conducted on four small-scale ECG databases, and the results demonstrate that the proposed method performs competitively with state-of-the-art methods.

Automatic text summarization (ATS) has achieved impressive performance thanks to recent advances in deep learning (DL) and the availability of large-scale corpora. The key points in ATS are to estimate the salience of information and to generate coherent results. Recently, a variety of DL-based approaches have been developed for better considering these two aspects. However, there is still a lack of comprehensive literature review for DL-based ATS approaches. The aim of this paper is to comprehensively review significant DL-based approaches that have been proposed in the literature with respect to the notion of generic ATS tasks and provide a walk-through of their evolution. We first give an overview of ATS and DL. The comparisons of the datasets are also given, which are commonly used for model training, validation, and evaluation. Then we summarize single-document summarization approaches. After that, an overview of multi-document summarization approaches is given. We further analyze the performance of the popular ATS models on common datasets. Various popular approaches can be employed for different ATS tasks. Finally, we propose potential research directions in this fast-growing field. We hope this exploration can provide new insights into future research of DL-based ATS.

This paper presents a survey of technologies for personal data self-management interfacing with administrative and territorial public service providers. It classifies a selection of scientific technologies into four categories of solutions:Personal Data Store (PDS), Identity Manager (IdM), Anonymous Certificate System and Access Control Delegation Architecture. Each category, along with its technological approach, is analyzed thanks to 18 identified functional criteria that encompass architectural and communication aspects, as well as user data lifecycle considerations. The originality of the survey is multifold. First, as far as we know, there is no such thorough survey covering such a panel of a dozen of existing solutions. Second, it is the first survey addressing Personally Identifiable Information (PII) management for both administrative and private service providers. Third, this paper achieves a functional comparison of solutions of very different technical natures. The outcome of this paper is the clear identification of functional gaps of each solution. As a result, this paper establishes the research directions to follow in order to fill these functional gaps.

Massive data is written to blockchain systems for the destination of keeping safe. However, existing blockchain protocols still demand that each full node has to contain the entire chain. Most nodes quit because they are unable to grow their storage space with the size of data. As the number of nodes decreases, the security of blockchains would significantly reduce. We present SE-Chain, a novel scale-out blockchain model that improves storage scalability under the premise of ensuring safety and achieves efficient retrieval. The SE-Chain consists of three parts:the data layer, the processing layer and the storage layer. In the data layer, each transaction is stored in the AB-M tree (Adaptive Balanced Merkle tree), which adaptively combines the advantages of balanced binary tree (quick retrieval) and Merkle tree (quick verification). In the processing layer, the full nodes store the part of the complete chain selected by the duplicate ratio regulation algorithm. Meanwhile, the node reliability verification method is used for increasing the stability of full nodes and reducing the risk of imperfect data recovering caused by the reduction of duplicate number in the storage layer. The experimental results on real datasets show that the query time of SE-Chain based on the AB-M tree is reduced by 17% when 16 nodes exist. Overall, SE-Chain improves the storage scalability extremely and implements efficient querying of transactions.

In online dynamic graph drawing, constraints over nodes and node pairs help preserve a coherent mental map in a sequence of graphs. Defining the constraints is challenging due to the requirements of both preserving mental map and satisfying the visual aesthetics of a graph layout. Most existing algorithms basically depend on local changes but fail to do proper evaluations on the global propagation when setting constraints. To solve this problem, we introduce a heuristic model derived from PageRank which simulates the node movement as an inverse Markov process hence to give a global analysis of the layout's change, according to which different constraints can be set. These constraints, along with stress function, generate layouts maintaining spatial positions and shapes of relatively stable substructures between adjacent graphs. Experiments demonstrate that our method preserves both structure and position similarity to help users track graph changes visually.