arXiv subjects

Qiang Yu

Publications and source records attributed to Qiang Yu.

24 records · Page 2Linked to original sources

arXiv2020

Constructing Accurate and Efficient Deep Spiking Neural Networks with Double-threshold and Augmented Schemes

Spiking neural networks (SNNs) are considered as a potential candidate to overcome current challenges such as the high-power consumption encountered by artificial neural networks (ANNs), however there is still a gap between them with respect to the recognition accuracy on practical tasks. A conversion strategy was thus introduced recently to bridge this gap by mapping a trained ANN to an SNN. However, it is still unclear that to what extent this obtained SNN can benefit both the accuracy advantage from ANN and high efficiency from the spike-based paradigm of computation. In this paper, we propose two new conversion methods, namely TerMapping and AugMapping. The TerMapping is a straightforward extension of a typical threshold-balancing method with a double-threshold scheme, while the AugMapping additionally incorporates a new scheme of augmented spike that employs a spike coefficient to carry the number of typical all-or-nothing spikes occurring at a time step. We examine the performance of our methods based on MNIST, Fashion-MNIST and CIFAR10 datasets. The results show that the proposed double-threshold scheme can effectively improve accuracies of the converted SNNs. More importantly, the proposed AugMapping is more advantageous for constructing accurate, fast and efficient deep SNNs as compared to other state-of-the-art approaches. Our study therefore provides new approaches for further integration of advanced techniques in ANNs to improve the performance of SNNs, which could be of great merit to applied developments with spike-based neuromorphic computing.

cs.NE↗

arXiv2020

Towards Efficient Processing and Learning with Spikes: New Approaches for Multi-Spike Learning

Spikes are the currency in central nervous systems for information transmission and processing. They are also believed to play an essential role in low-power consumption of the biological systems, whose efficiency attracts increasing attentions to the field of neuromorphic computing. However, efficient processing and learning of discrete spikes still remains as a challenging problem. In this paper, we make our contributions towards this direction. A simplified spiking neuron model is firstly introduced with effects of both synaptic input and firing output on membrane potential being modeled with an impulse function. An event-driven scheme is then presented to further improve the processing efficiency. Based on the neuron model, we propose two new multi-spike learning rules which demonstrate better performance over other baselines on various tasks including association, classification, feature detection. In addition to efficiency, our learning rules demonstrate a high robustness against strong noise of different types. They can also be generalized to different spike coding schemes for the classification task, and notably single neuron is capable of solving multi-category classifications with our learning rules. In the feature detection task, we re-examine the ability of unsupervised STDP with its limitations being presented, and find a new phenomenon of losing selectivity. In contrast, our proposed learning rules can reliably solve the task over a wide range of conditions without specific constraints being applied. Moreover, our rules can not only detect features but also discriminate them. The improved performance of our methods would contribute to neuromorphic computing as a preferable choice.

cs.NE↗

arXiv2019

Robust Environmental Sound Recognition with Sparse Key-point Encoding and Efficient Multi-spike Learning

The capability for environmental sound recognition (ESR) can determine the fitness of individuals in a way to avoid dangers or pursue opportunities when critical sound events occur. It still remains mysterious about the fundamental principles of biological systems that result in such a remarkable ability. Additionally, the practical importance of ESR has attracted an increasing amount of research attention, but the chaotic and non-stationary difficulties continue to make it a challenging task. In this study, we propose a spike-based framework from a more brain-like perspective for the ESR task. Our framework is a unifying system with a consistent integration of three major functional parts which are sparse encoding, efficient learning and robust readout. We first introduce a simple sparse encoding where key-points are used for feature representation, and demonstrate its generalization to both spike and non-spike based systems. Then, we evaluate the learning properties of different learning rules in details with our contributions being added for improvements. Our results highlight the advantages of the multi-spike learning, providing a selection reference for various spike-based developments. Finally, we combine the multi-spike readout with the other parts to form a system for ESR. Experimental results show that our framework performs the best as compared to other baseline approaches. In addition, we show that our spike-based framework has several advantageous characteristics including early decision making, small dataset acquiring and ongoing dynamic processing. Our framework is the first attempt to apply the multi-spike characteristic of nervous neurons to ESR. The outstanding performance of our approach would potentially contribute to draw more research efforts to push the boundaries of spike-based paradigm to a new horizon.

cs.NE↗

arXiv2013

Inverse Photoelectrochemical Cell

The splitting of water with sunlight using photoelectrochemical cell (PEC) to produce hydrogen is a promising avenue for sustainable energy production. The greatest virtue of PEC is that it uses sunlight as the only source to split water, but its efficiencies are still quite low due to poor performances of the available materials (such as SrTiO3). Consequently, another way of PEC research has been developed. By simultaneously using sunlight and electricity as energy source, PEC can get a larger current at a lower voltage, i.e., hydrogen can be made under the voltage below 1.23V, the minimum voltage required by electrolysis of water. But so far the efficiencies of the mainstream materials (such as TiO2) remain low. Where is the future development direction of PEC? Here we propose a new PEC model. Its operating principle is quite the opposite of the aforesaid conventional PEC, that is, the previous photoanode/photocathode is converted into the present photocathode/photoanode. It can also obtain high current under low voltage, even near zero voltage in extreme conditions. A basic single configuration and an improved n-p configuration were designed. We reselected materials for photoelectrodes, and confirmed its feasibility successfully. Two preliminary results were obtained from two contrasts: whether compared with water electrolysis or conventional PEC water splitting, the inverse PEC showed promising superiority.

physics.chem-ph↗

arXiv2008

New Reliability Assessment Method for Solder Joints in BGA Package by Considering the Interaction between Design Factors

As the integration and the miniaturization of electronics devices, design space become narrower and interactions between design factors affect their reliability. This paper presents a methodology of quantifying the interaction of each design factor in electronics devices. Thermal fatigue reliability of BGA assembly was assessed with the consideration of the interaction between design factors. Sensitivity analysis shows the influence of each design factor to inelastic strain range of a solder joint characterizing the thermal fatigue life if no interaction occurs. However, there is the interaction in BGA assembly since inelastic strain range depends on not only a mismatch in CTE but also a warpage of components. Clustering can help engineers to clarify the relation between design factors. The variation in the influence was taken to quantify the interaction of each design factor. Based on the interaction, simple evaluating approach of inelastic strain range for the BGA assembly was also developed. BGA package was simplified into a homogeneous component and equivalent CTE wascalculated from the warpage of BGA and PCB. The estimated equation was derived by using the response surface method as a function of design factors. Based upon these analytical results, design engineers can rate each factor's effect on reliability and assess the reliability of their basic design plan at the concept design stage.

physics.gen-ph↗

arXiv2007

Reliability Evaluation Method for Electronic Device BGA Package Considering the Interaction Between Design Factors

The recent development of electric and electronic devices has been remarkable. The miniaturization of electronic devices and high integration are progressing by advances in mounting technology. As a result, the reliability of fatigue life has been prioritized as an important concern, since the thermal expansion difference between a package and printed circuit board causes thermal fatigue. It is demanded a long-life product which has short development time. However, it is difficult because of interaction between each design factor. The authors have investigated the influence of various design factors on the reliability of soldered joints in BGA model by using response surface method and cluster analysis. By using these techniques, the interaction of all design factors was clarified. Based upon the analytical results, design engineers can rate each factor's effect on reliability and assess the reliability of their basic design plan at the concept design stage.

cond-mat.mtrl-sci↗