Showing papers by "Hong Kong University of Science and Technology published in 2015"

Aggregation-Induced Emission: Together We Shine, United We Soar!

Abstract: Ju Mei,†,‡,∥ Nelson L. C. Leung,†,‡,∥ Ryan T. K. Kwok,†,‡ Jacky W. Y. Lam,†,‡ and Ben Zhong Tang*,†,‡,§ †HKUST-Shenzhen Research Institute, Hi-Tech Park, Nanshan, Shenzhen 518057, China ‡Department of Chemistry, HKUST Jockey Club Institute for Advanced Study, Institute of Molecular Functional Materials, Division of Biomedical Engineering, State Key Laboratory of Molecular Neuroscience, Division of Life Science, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China Guangdong Innovative Research Team, SCUT-HKUST Joint Research Laboratory, State Key Laboratory of Luminescent Materials and Devices, South China University of Technology, Guangzhou 510640, China

Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting

Abstract: The goal of precipitation nowcasting is to predict the future rainfall intensity in a local region over a relatively short period of time. Very few previous studies have examined this crucial and challenging weather forecasting problem from the machine learning perspective. In this paper, we formulate precipitation nowcasting as a spatiotemporal sequence forecasting problem in which both the input and the prediction target are spatiotemporal sequences. By extending the fully connected LSTM (FC-LSTM) to have convolutional structures in both the input-to-state and state-to-state transitions, we propose the convolutional LSTM (ConvLSTM) and use it to build an end-to-end trainable model for the precipitation nowcasting problem. Experiments show that our ConvLSTM network captures spatiotemporal correlations better and consistently outperforms FC-LSTM and the state-of-the-art operational ROVER algorithm for precipitation nowcasting.

Neuroinflammation in Alzheimer's disease

Abstract: Increasing evidence suggests that Alzheimer's disease pathogenesis is not restricted to the neuronal compartment, but includes strong interactions with immunological mechanisms in the brain. Misfolded and aggregated proteins bind to pattern recognition receptors on microglia and astroglia, and trigger an innate immune response characterised by release of inflammatory mediators, which contribute to disease progression and severity. Genome-wide analysis suggests that several genes that increase the risk for sporadic Alzheimer's disease encode factors that regulate glial clearance of misfolded proteins and the inflammatory reaction. External factors, including systemic inflammation and obesity, are likely to interfere with immunological processes of the brain and further promote disease progression. Modulation of risk factors and targeting of these immune mechanisms could lead to future therapeutic or preventive strategies for Alzheimer's disease.

Convolutional LSTM Network: a machine learning approach for precipitation nowcasting

Abstract: The goal of precipitation nowcasting is to predict the future rainfall intensity in a local region over a relatively short period of time. Very few previous studies have examined this crucial and challenging weather forecasting problem from the machine learning perspective. In this paper, we formulate precipitation nowcasting as a spatiotemporal sequence forecasting problem in which both the input and the prediction target are spatiotemporal sequences. By extending the fully connected LSTM (FC-LSTM) to have convolutional structures in both the input-to-state and state-to-state transitions, we propose the convolutional LSTM (ConvLSTM) and use it to build an end-to-end trainable model for the precipitation nowcasting problem. Experiments show that our ConvLSTM network captures spatiotemporal correlations better and consistently outperforms FC-LSTM and the state-of-the-art operational ROVER algorithm for precipitation nowcasting.

Combined Measurement of the Higgs Boson Mass in pp Collisions at √s=7 and 8 TeV with the ATLAS and CMS Experiments

Abstract: A measurement of the Higgs boson mass is presented based on the combined data samples of the ATLAS and CMS experiments at the CERN LHC in the H→γγ and H→ZZ→4l decay channels. The results are obtained from a simultaneous fit to the reconstructed invariant mass peaks in the two channels and for the two experiments. The measured masses from the individual channels and the two experiments are found to be consistent among themselves. The combined measured mass of the Higgs boson is mH=125.09±0.21 (stat)±0.11 (syst) GeV.

Collaborative Deep Learning for Recommender Systems

Abstract: Collaborative filtering (CF) is a successful approach commonly used by many recommender systems. Conventional CF-based methods use the ratings given to items by users as the sole source of information for learning to make recommendation. However, the ratings are often very sparse in many applications, causing CF-based methods to degrade significantly in their recommendation performance. To address this sparsity problem, auxiliary information such as item content information may be utilized. Collaborative topic regression (CTR) is an appealing recent method taking this approach which tightly couples the two components that learn from two different sources of information. Nevertheless, the latent representation learned by CTR may not be very effective when the auxiliary information is very sparse. To address this problem, we generalize recently advances in deep learning from i.i.d. input to non-i.i.d. (CF-based) input and propose in this paper a hierarchical Bayesian model called collaborative deep learning (CDL), which jointly performs deep representation learning for the content information and collaborative filtering for the ratings (feedback) matrix. Extensive experiments on three real-world datasets from different domains show that CDL can significantly advance the state of the art.

Biosensing by Luminogens with Aggregation-induced Emission Characteristics

Abstract: Fluorescent biosensors are powerful analytical tools for studying biological events in living systems. Luminescent materials with aggregation-induced emission (AIE) attributes have attracted much research interest and have been identified as a novel class of luminogens to develop fluorescent turn-on biosensors with superior sensitivity. In this Tutorial Review, we present an overview of the AIE phenomenon and its mechanism. We summarize the structural design and working principle of AIE biosensors developed recently. Typical examples of AIE biosensors are presented.

Nonstoichiometric Oxides as Low-Cost and Highly-Efficient Oxygen Reduction/Evolution Catalysts for Low-Temperature Electrochemical Devices

Abstract: Reduction/Evolution Catalysts for Low-Temperature Electrochemical Devices Dengjie Chen,†,⊥,∇ Chi Chen,†,⊥ Zarah Medina Baiyee,† Zongping Shao,‡,§ and Francesco Ciucci*,†,∥ †Department of Mechanical and Aerospace Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China ‡State Key Laboratory of Materials-Oriented Chemical Engineering, College of Chemistry & Chemical Engineering, Nanjing Tech University, No. 5 Xin Mofan Road, Nanjing 210009, China Department of Chemical Engineering, Curtin University, Perth, Western Australia 6845, Australia Department of Chemical and Biomolecular Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China

Evidence for two-dimensional Ising superconductivity in gated MoS2

Abstract: The Zeeman effect, which is usually detrimental to superconductivity, can be strongly protective when an effective Zeeman field from intrinsic spin-orbit coupling locks the spins of Cooper pairs in a direction orthogonal to an external magnetic field. We performed magnetotransport experiments with ionic-gated molybdenum disulfide transistors, in which gating prepared individual superconducting states with different carrier dopings, and measured an in-plane critical field B(c2) far beyond the Pauli paramagnetic limit, consistent with Zeeman-protected superconductivity. The gating-enhanced B(c2) is more than an order of magnitude larger than it is in the bulk superconducting phases, where the effective Zeeman field is weakened by interlayer coupling. Our study provides experimental evidence of an Ising superconductor, in which spins of the pairing electrons are strongly pinned by an effective Zeeman field.

Specific light-up bioprobes based on AIEgen conjugates

Abstract: Driven by the high demand for sensitive and specific tools for optical sensing and imaging, bioprobes with various working mechanisms and advanced functionalities are flourishing at an incredible speed. Conventional fluorescent probes suffer from the notorious effect of aggregation-caused quenching that imposes limitation on their labelling efficiency or concentration to achieve desired sensitivity. The recently emerged fluorogens with an aggregation-induced emission (AIE) feature offer a timely remedy to tackle the challenge. Utilizing the unique properties of AIE fluorogens (AIEgens), specific light-up probes have been constructed through functionalization with recognition elements, showing advantages such as low background interference, a high signal to noise ratio and superior photostability with activatable therapeutic effects. In this tutorial review, we summarize the recent progress in the development of specific AIEgen-based light-up bioprobes. Through illustration of their operation mechanisms and application examples, we hope to provide guidelines for the design of more advanced AIE sensing and imaging platforms with high selectivity, great sensitivity and wide adaptability to a broad range of biomedical applications.

The case for rejecting the amyloid cascade hypothesis

Abstract: Alzheimer's disease (AD) is a biologically complex neurodegenerative dementia. Nearly 20 years ago, with the combination of observations from biochemistry, neuropathology and genetics, a compelling hypothesis known as the amyloid cascade hypothesis was formulated. The core of this hypothesis is that it is pathological accumulations of amyloid-β, a peptide fragment of a membrane protein called amyloid precursor protein, that act as the root cause of AD and initiate its pathogenesis. Yet, with the passage of time, growing amounts of data have accumulated that are inconsistent with the basically linear structure of this hypothesis. And while there is fear in the field over the consequences of rejecting it outright, clinging to an inaccurate disease model is the option we should fear most. This Perspective explores the proposition that we are over-reliant on amyloid to define and diagnose AD and that the time has come to face our fears and reject the amyloid cascade hypothesis.

Metallic Iron–Nickel Sulfide Ultrathin Nanosheets As a Highly Active Electrocatalyst for Hydrogen Evolution Reaction in Acidic Media

Abstract: We report on the synthesis of iron-nickel sulfide (INS) ultrathin nanosheets by topotactic conversion from a hydroxide precursor The INS nanosheets exhibit excellent activity and stability in strong acidic solutions as a hydrogen evolution reaction (HER) catalyst, lending an attractive alternative to the Pt catalyst The metallic α-INS nanosheets show an even lower overpotential of 105 mV at 10 mA/cm2 and a smaller Tafel slope of 40 mV/dec With the help of DFT calculations, the high specific surface area, facile ion transport and charge transfer, abundant electrochemical active sites, suitable H+ adsorption, and H2 formation kinetics and energetics are proposed to contribute to the high activity of the INS ultrathin nanosheets toward HER

Hybrid halide perovskite solar cell precursors: colloidal chemistry and coordination engineering behind device processing for high efficiency.

Abstract: The precursor of solution-processed perovskite thin films is one of the most central components for high-efficiency perovskite solar cells. We first present the crucial colloidal chemistry visualization of the perovskite precursor solution based on analytical spectra and reveal that perovskite precursor solutions for solar cells are generally colloidal dispersions in a mother solution, with a colloidal size up to the mesoscale, rather than real solutions. The colloid is made of a soft coordination complex in the form of a lead polyhalide framework between organic and inorganic components and can be structurally tuned by the coordination degree, thereby primarily determining the basic film coverage and morphology of deposited thin films. By utilizing coordination engineering, particularly through employing additional methylammonium halide over the stoichiometric ratio for tuning the coordination degree and mode in the initial colloidal solution, along with a thermal leaching for the selective release of ex...

Geometric phase and band inversion in periodic acoustic systems

Abstract: The behaviour of sound waves in phononic crystals—metamaterials with spatially varying acoustic characteristics—is similar to that of electrons in solids. Now, phononic band inversion and Zak phases have been measured for a 1D phononic system.

Human Action Recognition Using Factorized Spatio-Temporal Convolutional Networks

Abstract: Human actions in video sequences are three-dimensional (3D) spatio-temporal signals characterizing both the visual appearance and motion dynamics of the involved humans and objects. Inspired by the success of convolutional neural networks (CNN) for image classification, recent attempts have been made to learn 3D CNNs for recognizing human actions in videos. However, partly due to the high complexity of training 3D convolution kernels and the need for large quantities of training videos, only limited success has been reported. This has triggered us to investigate in this paper a new deep architecture which can handle 3D signals more effectively. Specifically, we propose factorized spatio-temporal convolutional networks (FstCN) that factorize the original 3D convolution kernel learning as a sequential process of learning 2D spatial kernels in the lower layers (called spatial convolutional layers), followed by learning 1D temporal kernels in the upper layers (called temporal convolutional layers). We introduce a novel transformation and permutation operator to make factorization in FstCN possible. Moreover, to address the issue of sequence alignment, we propose an effective training and inference strategy based on sampling multiple video clips from a given action video sequence. We have tested FstCN on two commonly used benchmark datasets (UCF-101 and HMDB-51). Without using auxiliary training videos to boost the performance, FstCN outperforms existing CNN based methods and achieves comparable performance with a recent method that benefits from using auxiliary training videos.

High-quality sandwiched black phosphorus heterostructure and its quantum oscillations

Abstract: Black phosphorus is an atomically thin material that exhibits excellent properties for electronics applications, but these degrade in atmospheric conditions. Here, the authors demonstrate the fabrication of stable, ultra-clean and high-mobility black phosphorus sandwiched between the layers of boron nitride.

Aggregation-induced emission of siloles

Abstract: Aggregation-induced emission (AIE) is a unique and significant photophysical phenomenon that differs greatly from the commonly acknowledged aggregation-caused emission quenching observed for many π-conjugated planar chromophores. The mechanistic decipherment of the AIE phenomenon is of high importance for the advance of new AIE systems and exploitation of their potential applications. Propeller-like 2,3,4,5-tetraphenylsiloles are archetypal AIE-active luminogens, and have been adopted as a core part in the design of numerous luminescent materials with diverse functionalities. In this review article, we elucidate the impacts of substituents on the AIE activity and shed light on the structure–property relationship of siloles, with the aim of promoting the judicious design of AIE-active functional materials in the future. Recent representative advances of new silole-based functional materials and their potential applications are reviewed as well.

Achieving Persistent Room Temperature Phosphorescence and Remarkable Mechanochromism from Pure Organic Luminogens

Abstract: Persistent room temperature phosphorescence (RTP) from pure organic luminogens can be rationally realized based on the crystallization-induced phosphorescence phenomenon and severe crystallization. A perfect crystal with dense molecular packing and effective inter-molecular interactions isolates the triplet excitons from quenching sites and significantly blocks the high-energy vibrational dissipation, thus yielding long-lasting RTP.

Fronthaul-constrained cloud radio access networks: insights and challenges

Abstract: As a promising paradigm for fifth generation wireless communication systems, cloud radio access networks (C-RANs) have been shown to reduce both capital and operating expenditures, as well as to provide high spectral efficiency (SE) and energy efficiency (EE). The fronthaul in such networks, defined as the transmission link between the baseband unit and the remote radio head, requires a high capacity, but is often constrained. This article comprehensively surveys recent advances in fronthaul-constrained CRANs, including system architectures and key techniques. Particularly, major issues relating to the impact of the constrained fronthaul on SE/EE and quality of service for users, including compression and quantization, large-scale coordinated processing and clustering, and resource allocation optimization, are discussed together with corresponding potential solutions. Open issues in terms of software-defined networking, network function virtualization, and partial centralization are also identified.

Specific Light-Up Bioprobe with Aggregation-Induced Emission and Activatable Photoactivity for the Targeted and Image-Guided Photodynamic Ablation of Cancer Cells†

Abstract: Activatable photosensitizers (PSs) have been widely used for the simultaneous fluorescence imaging and photodynamic ablation of cancer cells. However, the ready aggregation of traditional PSs in aqueous media can lead to fluorescence quenching as well as reduced phototoxicity even in the activated form. We have developed a series of PSs that show aggregation-enhanced emission and phototoxicity and thus the exact opposite behavior to that of previously reported PSs. We further developed a dual-targeted enzyme-activatable bioprobe based on the optimized photosensitizer and describe simultaneous light-up fluorescence imaging and activated photodynamic therapy for specific cancer cells. The design of smart probes should thus open new opportunities for targeted and image-guided photodynamic therapy.

Optimal Denial-of-Service Attack Scheduling With Energy Constraint

Abstract: Security of Cyber-Physical Systems (CPS) has gained increasing attention in recent years. Most existing works mainly investigate the system performance given some attacking patterns. In this technical note, we investigate how an attacker should schedule its Denial-of-Service (DoS) attacks to degrade the system performance. Specifically, we consider the scenario where a sensor sends its data to a remote estimator through a wireless channel, while an energy-constrained attacker decides whether to jam the channel at each sampling time. We construct optimal attack schedules to maximize the expected average estimation error at the remote estimator. We also provide the optimal attack schedules when a special intrusion detection system (IDS) at the estimator is given. We further discuss the optimal attack schedules when the sensor has energy constraint. Numerical examples are presented to demonstrate the effectiveness of the proposed optimal attack schedules.

Recurrent filmwise and dropwise condensation on a beetle mimetic surface.

Abstract: Vapor condensation plays a key role in a wide range of industrial applications including power generation, thermal management, water harvesting and desalination. Fast droplet nucleation and efficient droplet departure as well as low interfacial thermal resistance are important factors that determine the thermal performances of condensation; however, these properties have conflicting requirements on the structural roughness and surface chemistry of the condensing surface or condensation modes (e.g., filmwise vs dropwise). Despite intensive efforts over the past few decades, almost all studies have focused on the dropwise condensation enabled by superhydrophobic surfaces. In this work, we report the development of a bioinspired hybrid surface with high wetting contrast that allows for seamless integration of filmwise and dropwise condensation modes. We show that the synergistic cooperation in the observed recurrent condensation modes leads to improvements in all aspects of heat transfer properties including droplet nucleation density, growth rate, and self-removal, as well as overall heat transfer coefficient. Moreover, we propose an analytical model to optimize the surface morphological features for dramatic heat transfer enhancement.

Search for new phenomena in final states with an energetic jet and large missing transverse momentum in pp collisions at √=s=8 TeV with the ATLAS detector

Abstract: Results of a search for new phenomena in final states with an energetic jet and large missing transverse momentum are reported. The search uses 20.3 fb(-1) of root s = 8 TeV data collected in 2012 ...

First-principles study of anisotropic thermoelectric transport properties of IV-VI semiconductor compounds SnSe and SnS

Abstract: Tin selenide (SnSe) and tin sulfide (SnS) have recently attracted particular interest due to their great potential for large-scale thermoelectric applications. A complete prediction of the thermoelectric performance and the understanding of underlying heat and charge transport details are the key to further improvement of their thermoelectric efficiency. We conduct comprehensive investigations of both thermal and electrical transport properties of SnSe and SnS using first-principles calculations combined with the Boltzmann transport theory. Due to the distinct layered lattice structure, SnSe and SnS exhibit similarly anisotropic thermal and electrical behaviors. The cross-plane lattice thermal conductivity ${\ensuremath{\kappa}}_{L}$ is $40--60%$ lower than the in-plane values. Extremely low ${\ensuremath{\kappa}}_{L}$ is found for both materials because of high anharmonicity, while the average ${\ensuremath{\kappa}}_{L}$ of SnS is $\ensuremath{\sim}8%$ higher than that of SnSe from 300 to 750 K. It is suggested that nanostructuring would be difficult to further decrease ${\ensuremath{\kappa}}_{L}$ because of the short mean free paths of dominant phonon modes (1--30 nm at 300 K), while alloying would be efficient in reducing ${\ensuremath{\kappa}}_{L}$ considering that the relative ${\ensuremath{\kappa}}_{L}$ contribution $(\ensuremath{\sim}65%)$ of optical phonons is remarkably large. On the electrical side, the anisotropic electrical conductivities are mainly due to the different effective masses of holes and electrons along the $a, b$, and $c$ axes. This leads to the highest optimal $\mathit{ZT}$ values along the $b$ axis and lowest ones along the $a$ axis in both $p$-type materials. However, the $n$-type ones exhibit the highest $\mathit{ZT}\mathrm{s}$ along the $a$ axis due to the enhancement of power factor when the chemical potential gradually approaches the secondary conduction band valley that causes significant increase in electron mobility and density of states. Owing to the larger mobility and smaller ${\ensuremath{\kappa}}_{L}$ along the given direction, SnSe exhibits larger optimal ZTs compared with SnS in both $p$- and $n$-type materials. For both materials, the peak $\mathit{ZT}\mathrm{s}$ of $n$-type materials are much higher than those of $p$-type ones along the same direction. The predicted highest $\mathit{ZT}$ values at 750 K are 1.0 in SnSe and 0.6 in SnS along the $b$ axis for the $p$-type doping, while those for the $n$-type doping reach 2.7 in SnSe and 1.5 in SnS along the $a$ axis, rendering them among the best bulk thermoelectric materials for large-scale applications. Our calculations show reasonable agreements with the experimental results and quantitatively predict the great potential in further enhancing the thermoelectric performance of SnSe and SnS, especially for the $n$-type materials.

Human Action Recognition using Factorized Spatio-Temporal Convolutional Networks

Abstract: Human actions in video sequences are three-dimensional (3D) spatio-temporal signals characterizing both the visual appearance and motion dynamics of the involved humans and objects. Inspired by the success of convolutional neural networks (CNN) for image classification, recent attempts have been made to learn 3D CNNs for recognizing human actions in videos. However, partly due to the high complexity of training 3D convolution kernels and the need for large quantities of training videos, only limited success has been reported. This has triggered us to investigate in this paper a new deep architecture which can handle 3D signals more effectively. Specifically, we propose factorized spatio-temporal convolutional networks (FstCN) that factorize the original 3D convolution kernel learning as a sequential process of learning 2D spatial kernels in the lower layers (called spatial convolutional layers), followed by learning 1D temporal kernels in the upper layers (called temporal convolutional layers). We introduce a novel transformation and permutation operator to make factorization in FstCN possible. Moreover, to address the issue of sequence alignment, we propose an effective training and inference strategy based on sampling multiple video clips from a given action video sequence. We have tested FstCN on two commonly used benchmark datasets (UCF-101 and HMDB-51). Without using auxiliary training videos to boost the performance, FstCN outperforms existing CNN based methods and achieves comparable performance with a recent method that benefits from using auxiliary training videos.

Personality Traits and Personal Values: A Meta-Analysis

Abstract: Personality traits and personal values are important psychological characteristics, serving as important predictors of many outcomes. Yet, they are frequently studied separately, leaving the field with a limited understanding of their relationships. We review existing perspectives regarding the nature of the relationships between traits and values and provide a conceptual underpinning for understanding the strength of these relationships. Using 60 studies, we present a meta-analysis of the relationships between the Five-Factor Model (FFM) of personality traits and the Schwartz values, and demonstrate consistent and theoretically meaningful relationships. However, these relationships were not generally large, demonstrating that traits and values are distinct constructs. We find support for our premise that more cognitively based traits are more strongly related to values and more emotionally based traits are less strongly related to values. Findings also suggest that controlling for personal scale-use tendencies in values is advisable.

DevNet: A Deep Event Network for multimedia event detection and evidence recounting

Abstract: In this paper, we focus on complex event detection in internet videos while also providing the key evidences of the detection results. Convolutional Neural Networks (CNNs) have achieved promising performance in image classification and action recognition tasks. However, it remains an open problem how to use CNNs for video event detection and recounting, mainly due to the complexity and diversity of video events. In this work, we propose a flexible deep CNN infrastructure, namely Deep Event Network (DevNet), that simultaneously detects pre-defined events and provides key spatial-temporal evidences. Taking key frames of videos as input, we first detect the event of interest at the video level by aggregating the CNN features of the key frames. The pieces of evidences which recount the detection results, are also automatically localized, both temporally and spatially. The challenge is that we only have video level labels, while the key evidences usually take place at the frame levels. Based on the intrinsic property of CNNs, we first generate a spatial-temporal saliency map by back passing through DevNet, which then can be used to find the key frames which are most indicative to the event, as well as to localize the specific spatial position, usually an object, in the frame of the highly indicative area. Experiments on the large scale TRECVID 2014 MEDTest dataset demonstrate the promising performance of our method, both for event detection and evidence recounting.

Terthiophene-Based D–A Polymer with an Asymmetric Arrangement of Alkyl Chains That Enables Efficient Polymer Solar Cells

Abstract: We report a series of difluorobenzothiadizole (ffBT) and oligothiophene-based polymers with the oligothiophene unit being quaterthiophene (T4), terthiophene (T3), and bithiophene (T2). We demonstrate that a polymer based on ffBT and T3 with an asymmetric arrangement of alkyl chains enables the fabrication of 10.7% efficiency thick-film polymer solar cells (PSCs) without using any processing additives. By decreasing the number of thiophene rings per repeating unit and thus increasing the effective density of the ffBT unit in the polymer backbone, the HOMO and LUMO levels of the T3 polymers are significantly deeper than those of the T4 polymers, and the absorption onset of the T3 polymers is also slightly red-shifted. For the three T3 polymers obtained, the positions and size of the alkyl chains play a critical role in achieving the best PSC performances. The T3 polymer with a commonly known arrangement of alkyl chains (alkyl chains sitting on the first and third thiophenes in a mirror symmetric manner) yie...