Beihang University

About: Beihang University is a education organization based out in Beijing, China. It is known for research contribution in the topics: Computer science & Control theory. The organization has 67002 authors who have published 73507 publications receiving 975691 citations. The organization is also known as: Beijing University of Aeronautics and Astronautics.

Ranking with recursive neural networks and its application to multi-document summarization

Abstract: We develop a Ranking framework upon Recursive Neural Networks (R2N2) to rank sentences for multi-document summarization. It formulates the sentence ranking task as a hierarchical regression process, which simultaneously measures the salience of a sentence and its constituents (e.g., phrases) in the parsing tree. This enables us to draw on word-level to sentence-level supervisions derived from reference summaries. In addition, recursive neural networks are used to automatically learn ranking features over the tree, with hand-crafted feature vectors of words as inputs. Hierarchical regressions are then conducted with learned features concatenating raw features. Ranking scores of sentences and words are utilized to effectively select informative and non-redundant sentences to generate summaries. Experiments on the DUC 2001, 2002 and 2004 multi-document summarization datasets show that R2N2 outperforms state-of-the-art extractive summarization approaches.

Alkyl Side-Chain Engineering in Wide-Bandgap Copolymers Leading to Power Conversion Efficiencies over 10.

Abstract: T.L. and X.P. contributed equally to this work. This work was financially supported by the National Natural Science Foundation of China (NSFC) (Grant Nos. 51273203, 51261160496, 51473009, 21504066, 21534003, and 21303167), the International Science a Technology Cooperation Program of China (Grant No. 2014DFA52820), the Ministry of Science and Technology (Grant No. 2016YFA0200700), and RGC-NSFC grant N_HKU709/12 from the Research Grants Council of Hong Kong Special Administrative Region, China. This work was also supported by the National Research Foundation of Korea (Grant No. NRF-2015R1C1A1A02036599). X-ray data was acquired at beamlines 7.3.3 and 11.0.1.2 at the Advanced Light Source, which was supported by the Director, Office of Science, Office of Basic Energy Sciences, of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231. The authors thank Chenhui Zhu at beamline 7.3.3, and Cheng Wang at beamline 11.0.1.2 for assistance with data acquisition. Y.S. gratefully acknowledge Dr. Feng Liu for fruitful discussions and valuable suggestions.

Multilevel Cloud Detection in Remote Sensing Images Based on Deep Learning

Abstract: Cloud detection is one of the important tasks for remote sensing image processing. In this paper, a novel multilevel cloud detection method based on deep learning is proposed for remote sensing images. First, the simple linear iterative clustering (SLIC) method is improved to segment the image into good quality superpixels. Then, a deep convolutional neural network (CNN) with two branches is designed to extract the multiscale features from each superpixel and predict the superpixel as one of three classes including thick cloud, thin cloud, and noncloud. Finally, the predictions of all the superpixels in the image yield the cloud detection result. In the proposed cloud detection framework, the improved SLIC method can obtain accurate cloud boundaries by optimizing initial cluster centers, designing dynamic distance measure, and expanding search space. Moreover, different from traditional cloud detection methods that cannot achieve multilevel detection of cloud, the designed deep CNN model can not only detect cloud but also distinguish thin cloud from thick cloud. Experimental results indicate that the proposed method can detect cloud with higher accuracy and robustness than compared methods.

Learning Compact and Discriminative Stacked Autoencoder for Hyperspectral Image Classification

Abstract: As one of the fundamental research topics in remote sensing image analysis, hyperspectral image (HSI) classification has been extensively studied so far. However, how to discriminatively learn a low-dimensional feature space, in which the mapped features have small within-class scatter and big between-class separation, is still a challenging problem. To address this issue, this paper proposes an effective framework, named compact and discriminative stacked autoencoder (CDSAE), for HSI classification. The proposed CDSAE framework comprises two stages with different optimization objectives, which can learn discriminative low-dimensional feature mappings and train an effective classifier progressively. First, we impose a local Fisher discriminant regularization on each hidden layer of stacked autoencoder (SAE) to train discriminative SAE (DSAE) by minimizing reconstruction error. This stage can learn feature mappings, in which the pixels from the same land-cover class are mapped as nearly as possible and the pixels from different land-cover categories are separated by a large margin. Second, we learn an effective classifier and meanwhile update DSAE with a local Fisher discriminant regularization being embedded on the top of feature representations. Moreover, to learn a compact DSAE with as small number of hidden neurons as possible, we impose a diversity regularization on the hidden neurons of DSAE to balance the feature dimensionality and the feature representation capability. The experimental results on three widely-used HSI data sets and comprehensive comparisons with existing methods demonstrate that our proposed method is effective.