arXiv subjects

Meng Han

Publications and source records attributed to Meng Han.

111 records · Page 7Linked to original sources

arXiv2019

Large-scale Multi-view Subspace Clustering in Linear Time

A plethora of multi-view subspace clustering (MVSC) methods have been proposed over the past few years. Researchers manage to boost clustering accuracy from different points of view. However, many state-of-the-art MVSC algorithms, typically have a quadratic or even cubic complexity, are inefficient and inherently difficult to apply at large scales. In the era of big data, the computational issue becomes critical. To fill this gap, we propose a large-scale MVSC (LMVSC) algorithm with linear order complexity. Inspired by the idea of anchor graph, we first learn a smaller graph for each view. Then, a novel approach is designed to integrate those graphs so that we can implement spectral clustering on a smaller graph. Interestingly, it turns out that our model also applies to single-view scenario. Extensive experiments on various large-scale benchmark data sets validate the effectiveness and efficiency of our approach with respect to state-of-the-art clustering methods.

cs.LG↗

arXiv2019

Adversarial Examples Versus Cloud-based Detectors: A Black-box Empirical Study

Deep learning has been broadly leveraged by major cloud providers, such as Google, AWS and Baidu, to offer various computer vision related services including image classification, object identification, illegal image detection, etc. While recent works extensively demonstrated that deep learning classification models are vulnerable to adversarial examples, cloud-based image detection models, which are more complicated than classifiers, may also have similar security concern but not get enough attention yet. In this paper, we mainly focus on the security issues of real-world cloud-based image detectors. Specifically, (1) based on effective semantic segmentation, we propose four attacks to generate semantics-aware adversarial examples via only interacting with black-box APIs; and (2) we make the first attempt to conduct an extensive empirical study of black-box attacks against real-world cloud-based image detectors. Through the comprehensive evaluations on five major cloud platforms: AWS, Azure, Google Cloud, Baidu Cloud, and Alibaba Cloud, we demonstrate that our image processing based attacks can reach a success rate of approximately 100%, and the semantic segmentation based attacks have a success rate over 90% among different detection services, such as violence, politician, and pornography detection. We also proposed several possible defense strategies for these security challenges in the real-life situation.

cs.CV↗

arXiv2016

Strongly Anisotropic Thermal and Electrical Conductivities of Self-assembled Silver Nanowire Network

Heat dissipation issues are the emerging challenges in the field of flexible electronics. Thermal management of flexible electronics creates a demand for flexible materials with highly anisotropic thermal conductivity, which work as heat spreaders to remove excess heat in the in-plane direction and as heat shields to protect human skin or device components under them from heating. This study proposes a self-assembled silver nanowire network with high thermal and electrical anisotropy with the potential to solve these challenges. The in-plane thermal conductivity of the network along the axial direction of silver nanowires is measured as 37 W/m-K while the cross-plane thermal conductivity is only 0.36 W/m-K. The results of measurements of electrical and thermal conductivities suggest that abundant wire-wire contacts strongly impede thermal transport. The excellent alignment of nanowires results in the same anisotropy ratio of 3 for both thermal and electrical conduction in the two in-plane directions. The ratio remains unchanged as the temperature decrease to 50 K, which indicates that wire-wire contacts lower the thermal and electrical conduction in the two directions to the same extent and their effect is independent of temperature. In addition, phonon softening markedly reduces the Debye temperatures of the network, which are fitted from the electrical resistivity data. As a result of phonon thermal conduction, the Lorenz numbers of the film in the two directions, which are approximately the same, are larger than the Sommerfeld value at room temperature and decrease as temperature decreases because of small angle scattering and the reduced phonon contribution. This nanowire network provides a solution to the emerging challenges of thermal management of flexible electronics.

cond-mat.mes-hall↗