Nanyang Technological University

About: Nanyang Technological University is a education organization based out in Singapore, Singapore. It is known for research contribution in the topics: Computer science & Catalysis. The organization has 48003 authors who have published 112815 publications receiving 3294199 citations. The organization is also known as: NTU & Universiti Teknologi Nanyang.

Shape-controlled synthesis of MnO2 nanostructures with enhanced electrocatalytic activity for oxygen reduction

Abstract: In this work, three types of MnO2 nanostructures, viz., microsphere/nanosheet core−corona hierarchical architectures, one-dimensional (1D) nanorods, and nanotubes, have been synthesized employing a simple hydrothermal process. The formation mechanisms have been rationalized. The materials have been thoroughly characterized by X-ray diffraction, Brunauer−Emmett−Teller spectrometry, field-emission scanning electron miscroscopy, energy dispersive spectroscopy, and transmission electron microscopy. The microsphere/nanosheet core−corona hierarchical structures are found to be the layered birnessite-type MnO2, while 1D nanorods and nanotubes are of the α-MnO2 phase. These MnO2 nanostructures are used as a model system for studying the shape/phase-dependent electrocatalytic properties for the oxygen reduction reaction, which have be investigated by cyclic and linear sweep voltammetry. It is found that α-MnO2 nanorods/tubes possess largely enhanced electrocatalytic activity compared to birnessite-type MnO2 core−c...

Coordination polymer structure and revisited hydrogen evolution catalytic mechanism for amorphous molybdenum sulfide

Abstract: Molybdenum sulfides are very attractive noble-metal-free electrocatalysts for the hydrogen evolution reaction (HER) from water. The atomic structure and identity of the catalytically active sites have been well established for crystalline molybdenum disulfide (c-MoS2) but not for amorphous molybdenum sulfide (a-MoSx), which exhibits significantly higher HER activity compared to its crystalline counterpart. Here we show that HER-active a-MoSx, prepared either as nanoparticles or as films, is a molecular-based coordination polymer consisting of discrete [Mo3S13]2- building blocks. Of the three terminal disulfide (S22-) ligands within these clusters, two are shared to form the polymer chain. The third one remains free and generates molybdenum hydride moieties as the active site under H2 evolution conditions. Such a molecular structure therefore provides a basis for revisiting the mechanism of a-MoSx catalytic activity, as well as explaining some of its special properties such as reductive activation and corrosion. Our findings open up new avenues for the rational optimization of this HER electrocatalyst as an alternative to platinum.

SLAYER: Spike Layer Error Reassignment in Time

Abstract: Configuring deep Spiking Neural Networks (SNNs) is an exciting research avenue for low power spike event based computation. However, the spike generation function is non-differentiable and therefore not directly compatible with the standard error backpropagation algorithm. In this paper, we introduce a new general backpropagation mechanism for learning synaptic weights and axonal delays which overcomes the problem of non-differentiability of the spike function and uses a temporal credit assignment policy for backpropagating error to preceding layers. We describe and release a GPU accelerated software implementation of our method which allows training both fully connected and convolutional neural network (CNN) architectures. Using our software, we compare our method against existing SNN based learning approaches and standard ANN to SNN conversion techniques and show that our method achieves state of the art performance for an SNN on the MNIST, NMNIST, DVS Gesture, and TIDIGITS datasets.

Gated Siamese Convolutional Neural Network Architecture for Human Re-Identification

Abstract: Matching pedestrians across multiple camera views, known as human re-identification, is a challenging research problem that has numerous applications in visual surveillance. With the resurgence of Convolutional Neural Networks (CNNs), several end-to-end deep Siamese CNN architectures have been proposed for human re-identification with the objective of projecting the images of similar pairs (i.e. same identity) to be closer to each other and those of dissimilar pairs to be distant from each other. However, current networks extract fixed representations for each image regardless of other images which are paired with it and the comparison with other images is done only at the final level. In this setting, the network is at risk of failing to extract finer local patterns that may be essential to distinguish positive pairs from hard negative pairs. In this paper, we propose a gating function to selectively emphasize such fine common local patterns by comparing the mid-level features across pairs of images. This produces flexible representations for the same image according to the images they are paired with. We conduct experiments on the CUHK03, Market-1501 and VIPeR datasets and demonstrate improved performance compared to a baseline Siamese CNN architecture.