Xiamen University

About: Xiamen University is a education organization based out in Amoy, Fujian, China. It is known for research contribution in the topics: Catalysis & Population. The organization has 50472 authors who have published 54480 publications receiving 1058239 citations. The organization is also known as: Amoy University & Xiàmén Dàxué.

Accelerating convolutional networks via global & dynamic filter pruning

Abstract: Accelerating convolutional neural networks has recently received ever-increasing research focus. Among various approaches proposed in the literature, filter pruning has been regarded as a promising solution, which is due to its advantage in significant speedup and memory reduction of both network model and intermediate feature maps. To this end, most approaches tend to prune filters in a layerwise fixed manner, which is incapable to dynamically recover the previously removed filter, as well as jointly optimize the pruned network across layers. In this paper, we propose a novel global & dynamic pruning (GDP) scheme to prune redundant filters for CNN acceleration. In particular, GDP first globally prunes the unsalient filters across all layers by proposing a global discriminative function based on prior knowledge of each filter. Second, it dynamically updates the filter saliency all over the pruned sparse network, and then recovers the mistakenly pruned filter, followed by a retraining phase to improve the model accuracy. Specially, we effectively solve the corresponding nonconvex optimization problem of the proposed GDP via stochastic gradient descent with greedy alternative updating. Extensive experiments show that the proposed approach achieves superior performance to accelerate several cutting-edge CNNs on the ILSVRC 2012 benchmark, comparing to the state-of-the-art filter pruning methods.

ECG Analysis Using Multiple Instance Learning for Myocardial Infarction Detection

Abstract: This paper presents a useful technique for totally automatic detection of myocardial infarction from patients' ECGs. Due to the large number of heartbeats constituting an ECG and the high cost of having all the heartbeats manually labeled, supervised learning techniques have achieved limited success in ECG classification. In this paper, we first discuss the rationale for applying multiple instance learning (MIL) to automated ECG classification and then propose a new MIL strategy called latent topic MIL, by which ECGs are mapped into a topic space defined by a number of topics identified over all the unlabeled training heartbeats and support vector machine is directly applied to the ECG-level topic vectors. Our experimental results on real ECG datasets from the PTB diagnostic database demonstrate that, compared with existing MIL and supervised learning algorithms, the proposed algorithm is able to automatically detect ECGs with myocardial ischemia without labeling any heartbeats. Moreover, it improves classification quality in terms of both sensitivity and specificity.

Sub-5 nm Ultra-Fine FeP Nanodots as Efficient Co-Catalysts Modified Porous g-C3N4 for Precious-Metal-Free Photocatalytic Hydrogen Evolution under Visible Light

Abstract: Sub-5 nm ultra-fine iron phosphide (FeP) nano-dots-modified porous graphitic carbon nitride (g-C3N4) heterojunction nanostructures are successfully prepared through the gas-phase phosphorization of Fe3O4/g-C3N4 nanocomposites. The incorporation of zero-dimensional (0D) ultra-small FeP nanodots co-catalysts not only effectively facilitate charge separation but also serve as reaction active sites for hydrogen (H2) evolution. Herein, the strongly coupled FeP/g-C3N4 hybrid systems are employed as precious-metal-free photocatalysts for H2 production under visible-light irradiation. The optimized FeP/g-C3N4 sample displays a maximum H2 evolution rate of 177.9 μmol h-1 g-1 with the apparent quantum yield of 1.57% at 420 nm. Furthermore, the mechanism of photocatalytic H2 evolution using 0D/2D FeP/g-C3N4 heterojunction interfaces is systematically corroborated by steady-state photoluminescence (PL), time-resolved PL spectroscopy, and photoelectrochemical results. Additionally, an increased donor density in FeP/g-C3N4 is evidenced from the Mott-Schottky analysis in comparison with that of parent g-C3N4, signifying the enhancement of electrical conductivity and charge transport owing to the emerging role of FeP. The density functional theory calculations reveal that the FeP/g-C3N4 hybrids could act as a promising catalyst for the H2 evolution reaction. Overall, this work not only paves a new path in the engineering of monodispersed FeP-decorated g-C3N4 0D/2D robust nanoarchitectures but also elucidates potential insights for the utilization of noble-metal-free FeP nanodots as remarkable co-catalysts for superior photocatalytic H2 evolution.