Search

Exabytes of data are generated daily by humans, leading to the growing needs for new efforts in dealing with the grand challenges for multi-label learning brought by big data. For example, extreme multi-label classification is an active and rapidly growing research area that deals with classification tasks with extremely large number of classes or labels; utilizing massive data with limited supervision to build a multi-label classification model becomes valuable for practical applications, etc. Besides these, there are tremendous efforts on how to harvest the strong learning capability of deep learning to better capture the label dependencies in multi-label learning, which is the key for deep learning to address real-world classification tasks. However, it is noted that there have been a lack of systemic studies that focus explicitly on analyzing the emerging trends and new challenges of multi-label learning in the era of big data. It is imperative to call for a comprehensive survey to fulfil this mission and delineate future research directions and new applications. [ABSTRACT FROM AUTHOR]

Image scene classification aiming to assign specific semantic labels for each image is vitally important for the applications of remote sensing (RS) data. In real world, since the observation environment is open and dynamic, RS images are collected sequentially and the numbers of images and classes grow rapidly over time. Most existing scene classification methods are offline learning algorithms, which are inefficient and unscalable for this scenario. In this article, an incremental learning with open-set recognition (ILOSR) framework is proposed for RS image scene classification in the open and dynamic environment, which can identify the unknown classes from a stream of data and learn these new classes incrementally. Specifically, a controllable convex hull-based exemplar selection strategy is designed to address the catastrophic forgetting issue in incremental learning, which can reduce training time and memory footprint effectively. In addition, a new loss function based on prototype learning and uncertainty measurement is proposed for OSR to enhance the interclass discrimination and intraclass compactness of the learned deep features. Experimental results on real RS datasets demonstrate that the proposed method can not only outperform the state-of-the-art approaches on offline classification, incremental learning, and OSR problem separately but also achieve better and more stable performance in the experiments for ILOSR. [ABSTRACT FROM AUTHOR]

This article presents a simple and intuitive solution for multilabel image classification, which achieves the competitive performance on the popular COCO and PASCAL VOC benchmarks. The main idea is to capture how humans perform this task: we recognize both labels (i.e., objects and attributes) and the correlation of labels at the same time. Here, label recognition is performed by a standard ConvNet pipeline, whereas label correlation modeling is done by projecting both labels and image features extracted by the ConvNet to a common latent vector space. Specifically, we carefully design the loss function to ensure that: 1) labels and features that co-appear frequently are close to each other in the latent space and 2) conversely, labels/features that do not appear together are far apart. This information is then combined with the original ConvNet outputs to form the final prediction. The whole model is trained end-to-end, with no additional supervised information other than the image-level supervised information. Experiments show that the proposed method consistently outperforms previous approaches on COCO and PASCAL VOC in terms of mAP, macro/micro precision, recall, and $F$ -measure. Further, our model is highly efficient at test time, with only a small number of additional weights compared to the base model for direct label recognition. [ABSTRACT FROM AUTHOR]

Given two video frames $X_{0}$ and $X_{n+1}$ , we aim to generate a series of intermediate frames $Y_{1}, Y_{2}, \ldots, Y_{n}$ , such that the resulting video consisting of frames $X_{0}, Y_{1}-Y_{n}, and X_{n+1}$ appears realistic to a human watcher. Such video generation has numerous important applications, including video compression, movie production, slow-motion filming, video surveillance, and forensic analysis. Yet, video generation is highly challenging due to the vast search space of possible frames. Previous methods, mostly based on video prediction and/or video interpolation, tend to generate poor-quality videos with severe motion blur. This paper proposes a novel, end-to-end approach to video generation using generative adversarial networks (GANs). In particular, our design involves two concatenated GANs, one capturing motions and the other generating frame details. The loss function is also carefully engineered to include adversarial loss, gradient difference (for motion learning), and normalized product correlation loss (for frame details). Experiments using three video datasets, namely, Google Robotic Push, KTH human actions, and UCF101, demonstrate that the proposed solution generates high-quality, realistic, and sharp videos, whereas all previous solutions output noisy and blurry results. [ABSTRACT FROM AUTHOR]

Multi-output learning with the task of simultaneously predicting multiple outputs for an input has increasingly attracted interest from researchers due to its wide application. The $k$ nearest neighbor ($k \text{NN}$ ) algorithm is one of the most popular frameworks for handling multi-output problems. The performance of $k \text{NN}$ depends crucially on the metric used to compute the distance between different instances. However, our experiment results show that the existing advanced metric learning technique cannot provide an appropriate distance metric for multi-output tasks. This paper systematically studies how to efficiently learn an appropriate distance metric for multi-output problems with provable guarantee. In particular, we present a novel large margin metric learning paradigm for multi-output tasks, which projects both the input and output into the same embedding space and then learns a distance metric to discover output dependency such that instances with very different multiple outputs will be moved far away. Several strategies are then proposed to speed up the training and testing time. Moreover, we study the generalization error bound of our method for three learning tasks, which shows that our method converges to the optimal solutions. Experiments on three multi-output learning tasks (multi-label classification, multi-target regression, and multi-concept retrieval) validate the effectiveness and scalability of the proposed method. [ABSTRACT FROM AUTHOR]

Hyperspectral imagery (HSI) has shown promising results in real-world applications. However, the technological evolution of optical sensors poses two main challenges in HSI classification: 1) the spectral band is usually redundant and noisy and 2) HSI with millions of pixels has become increasingly common in real-world applications. Motivated by the recent success of hybrid huberized support vector machines (HHSVMs), which inherit the benefits of both lasso and ridge regression, this paper first investigates the advantages of HHSVM for HSI applications. Unfortunately, the existing HHSVM solvers suffer from prohibitive computational costs on large-scale data sets. To solve this problem, this paper proposes simple and effective stochastic HHSVM algorithms for HSI classification. In the stochastic settings, we show that with a probability of at least $1-\varrho $ , our algorithms find an $\epsilon $ -accurate solution using $\tilde {O}({1}/{\lambda _{2}\epsilon })$ iterations. Since the convergence rate of our algorithms does not depend on the size of the training set, our algorithms are suitable for handling large-scale problems. We demonstrate the superiority of our algorithms by conducting experiments on large-scale binary and multiclass classification problems, comparing to the state-of-the-art HHSVM solvers. Finally, we apply our algorithms to real HSI classification and achieve promising results. [ABSTRACT FROM AUTHOR]

In social networks, nodes (or users) interested in specific topics are often influenced by others. The influence is usually associated with a set of nodes rather than a single one. An interesting but challenging task for any given topic and node is to find the set of nodes that represents the source or trigger for the topic and thus identify those nodes that have the greatest influence on the given node as the topic spreads. We find that it is an NP-hard problem. This paper proposes an effective framework to deal with this problem. First, the topic propagation is represented as the Bayesian network. We then construct the propagation model by a variant of the voter model. The probability transition matrix (PTM) algorithm is presented to conduct the probability inference with the complexity \boldsymbol \mathcal O(\boldsymbol \theta ^\mathbf 3\mathbf log\mathbf {2}{\boldsymbol \theta }) , while \boldsymbol \theta is the number nodes in the given graph. To evaluate the PTM algorithm, we conduct extensive experiments on real datasets. The experimental results show that the PTM algorithm is both effective and efficient. [ABSTRACT FROM PUBLISHER]

The high-field quasi-continuous magnets (QCM) are of great interest for scientific experiments that require long pulse duration and in particular a field pulse with a flat-top. For the QCM, the power supply is a real challenge. At the Wuhan National High Magnetic Field Center (WHMFC), it is planned to realize a 45 T/500 ms flat-top field by the series combination of a 0.6 kV/35 kA battery bank and a 12-pulse rectifier that is powered by a 100 MVA/100 MJ flywheel pulse generator (FPG). In order to reduce the dc current ripple, a dc passive filter is connected in parallel with the rectifier. The simulation of the pulse shape with respect to the voltage profile supplied by the power supply takes the resistance of the magnet into account; this increases with temperature as a function of coil current. The configuration and the associated control strategy are presented in this paper. A simulation model of the combined power supply system is developed and compared with the model of a single rectifier supply system. Results are provided to validate the proposed method. [ABSTRACT FROM PUBLISHER]

At the Wuhan National High Magnetic Field Center (WHMFC) a 50 T/100 ms flat-top long-pulse magnet consisting of two coaxially nested coils is designed. The power and energy to operate the magnet is provided by a 100 MVA/185 MJ inertial energy storage motor-generator set. To accommodate the needs of this and future magnets, two 67.5 MW power supply modules based on 12-pulse converters are designed to convert the ac-power produced by the generator into controlled dc-power. The modules can operate independently as well as in parallel or in series. Each module consists of a combined input disconnect/grounding switch, two rectifier transformers, two parallel-connected 6-pulse rectifiers, a crowbar circuit, and a reversing switch. Two different transformer tap settings allow the module to provide no-load voltages of 3.2 kV and 2.7 kV and full-load voltages of 2.7 kV and 2.25 kV at the rated current of 25 kA for a 2 second pulse once every half hour. All components of the power supply modules are now under construction, with complete delivery and installation expected in early 2012. The design of the power supply components is described in detail. Digital simulation results of the power supply driving the 50 T/100 ms magnet are presented. [ABSTRACT FROM PUBLISHER]