arXiv subjects

Qing Wu

Publications and source records attributed to Qing Wu.

48 records · Page 3Linked to original sources

arXiv2020

Real time multibody modeling and simulation of a scaled bogie test rig

In wheel rail adhesion studies, most of the test rigs used are simplified designs such as a single wheel or wheelset, but the results may not be accurate. Alternatively, representing the complex system by using a full vehicle model provides accurate results but may incur complexity in design. To trade off accuracy over complexity, a bogie model can be the optimum selection. Furthermore, only a real time model can replicate its physical counterpart in the time domain. Developing such a model requires broad expertise and appropriate software and hardware. A few published works are available which deal with real time modeling. However, the influence of the control system has not been included in those works. To address these issues, a real-time scaled bogie test rig including the control system is essential. Therefore, a 1:4 scaled bogie roller rig is developed to study the adhesion between wheel and roller contact. To compare the performances obtained from the scaled bogie test rig and to expand the test applications, a numerical simulation model of that scaled bogie test rig is developed using Gensys multibody software. This model is the complete model of the test rig which delivers more precise results. To exactly represent the physical counterpart system in the time domain, a real-time scaled bogie test rig (RT SBTR) is developed after four consecutive stages. Then, to simulate the RT-SBTR to solve the internal state equations and functions representing the physical counterpart system in equal or less than actual time, the real-time simulation environment is prepared in two stages. To such end, the computational time improved from 4 times slower than real time to 2 times faster than real time. Finally, the real time scaled bogie model is also incorporated with the braking control system which slightly reduces the computational performances without affecting real time capability.

eess.SY↗

arXiv2020

Wheel-Rail Interface Condition Estimation (W-RICE)

The surface roughness between the wheel and rail has a huge influence on rolling noise level. The presence of the third body such as frost or grease at wheel-rail interface contributes towards change in adhesion coefficient resulting in the generation of noise at various levels. Therefore, it is possible to estimate adhesion conditions between the wheel and rail from the analysis of noise patterns originating from wheel-rail interaction. In this study, a new approach to estimate adhesion condition is proposed which takes rolling noise as input.

eess.AS↗

arXiv2020

Cross-Channel Intragroup Sparsity Neural Network

Modern deep neural networks rely on overparameterization to achieve state-of-the-art generalization. But overparameterized models are computationally expensive. Network pruning is often employed to obtain less demanding models for deployment. Fine-grained pruning removes individual weights in parameter tensors and can achieve a high model compression ratio with little accuracy degradation. However, it introduces irregularity into the computing dataflow and often does not yield improved model inference efficiency in practice. Coarse-grained model pruning, while realizing satisfactory inference speedup through removal of network weights in groups, e.g. an entire filter, often lead to significant accuracy degradation. This work introduces the cross-channel intragroup (CCI) sparsity structure, which can prevent the inference inefficiency of fine-grained pruning while maintaining outstanding model performance. We then present a novel training algorithm designed to perform well under the constraint imposed by the CCI-Sparsity. Through a series of comparative experiments we show that our proposed CCI-Sparsity structure and the corresponding pruning algorithm outperform prior art in inference efficiency by a substantial margin given suited hardware acceleration in the future.

cs.LG↗

arXiv2019

Adabot: Fault-Tolerant Java Decompiler

Reverse Engineering(RE) has been a fundamental task in software engineering. However, most of the traditional Java reverse engineering tools are strictly rule defined, thus are not fault-tolerant, which pose serious problem when noise and interference were introduced into the system. In this paper, we view reverse engineering as a statistical machine translation task instead of rule-based task, and propose a fault-tolerant Java decompiler based on machine translation models. Our model is based on attention-based Neural Machine Translation (NMT) and Transformer architectures. First, we measure the translation quality on both the redundant and purified datasets. Next, we evaluate the fault-tolerance(anti-noise ability) of our framework on test sets with different unit error probability (UEP). In addition, we compare the suitability of different word segmentation algorithms for decompilation task. Experimental results demonstrate that our model is more robust and fault-tolerant compared to traditional Abstract Syntax Tree (AST) based decompilers. Specifically, in terms of BLEU-4 and Word Error Rate (WER), our performance has reached 94.50% and 2.65% on the redundant test set; 92.30% and 3.48% on the purified test set.

cs.SE↗

arXiv2018

Shape-from-Mask: A Deep Learning Based Human Body Shape Reconstruction from Binary Mask Images

3D content creation is referred to as one of the most fundamental tasks of computer graphics. And many 3D modeling algorithms from 2D images or curves have been developed over the past several decades. Designers are allowed to align some conceptual images or sketch some suggestive curves, from front, side, and top views, and then use them as references in constructing a 3D model automatically or manually. However, to the best of our knowledge, no studies have investigated on 3D human body reconstruction in a similar manner. In this paper, we propose a deep learning based reconstruction of 3D human body shape from 2D orthographic views. A novel CNN-based regression network, with two branches corresponding to frontal and lateral views respectively, is designed for estimating 3D human body shape from 2D mask images. We train our networks separately to decouple the feature descriptors which encode the body parameters from different views, and fuse them to estimate an accurate human body shape. In addition, to overcome the shortage of training data required for this purpose, we propose some significantly data augmentation schemes for 3D human body shapes, which can be used to promote further research on this topic. Extensive experimen- tal results demonstrate that visually realistic and accurate reconstructions can be achieved effectively using our algorithm. Requiring only binary mask images, our method can help users create their own digital avatars quickly, and also make it easy to create digital human body for 3D game, virtual reality, online fashion shopping.

cs.GR↗

arXiv2018

Long short-term memory networks in memristor crossbars

Recent breakthroughs in recurrent deep neural networks with long short-term memory (LSTM) units has led to major advances in artificial intelligence. State-of-the-art LSTM models with significantly increased complexity and a large number of parameters, however, have a bottleneck in computing power resulting from limited memory capacity and data communication bandwidth. Here we demonstrate experimentally that LSTM can be implemented with a memristor crossbar, which has a small circuit footprint to store a large number of parameters and in-memory computing capability that circumvents the 'von Neumann bottleneck'. We illustrate the capability of our system by solving real-world problems in regression and classification, which shows that memristor LSTM is a promising low-power and low-latency hardware platform for edge inference.

cs.ET↗

arXiv2018

MAT: A Multi-strength Adversarial Training Method to Mitigate Adversarial Attacks

Some recent works revealed that deep neural networks (DNNs) are vulnerable to so-called adversarial attacks where input examples are intentionally perturbed to fool DNNs. In this work, we revisit the DNN training process that includes adversarial examples into the training dataset so as to improve DNN's resilience to adversarial attacks, namely, adversarial training. Our experiments show that different adversarial strengths, i.e., perturbation levels of adversarial examples, have different working zones to resist the attack. Based on the observation, we propose a multi-strength adversarial training method (MAT) that combines the adversarial training examples with different adversarial strengths to defend adversarial attacks. Two training structures - mixed MAT and parallel MAT - are developed to facilitate the tradeoffs between training time and memory occupation. Our results show that MAT can substantially minimize the accuracy degradation of deep learning systems to adversarial attacks on MNIST, CIFAR-10, CIFAR-100, and SVHN.

cs.LG↗

arXiv2018

Observation of unusual optical band structure of CH3NH3PbI3 perovskite single crystal

Extensive efforts have been undertaken on the photoelectric physics of hybrid organolead halide perovskites to unveil the reason for the attractive photovoltaic performance. Yet, the resulting evidences are far from being fully conclusive. Herein, we provide another direct support for this issue. In addition to the observation on the conventional band edge at 1.58 eV that presents a blueshift toward temperature increase, interestingly, we also observe an unusual optical band edge at 1.48 eV in CH3NH3PbI3 perovskite single crystals. Contrary to the conventional band edge, this one shows an obvious redshift toward the enhancement in temperature, in agreement with the Varshni relation. More interestingly, the unusual band edge exhibits a series of obvious absorption and photocurrent signals, but the according photoluminescence signals are not observable. This indicates that this band edge is particularly beneficial for the photovoltaic effect due to the inhibited radiative recombination. The kinetics on photo-involved charge transition and transfer are investigated using the pump-probe photoconductivity technique, and a changeable band structure model was proposed.

cond-mat.mtrl-sci↗

arXiv2017

Testing as an Investment

Software testing is an expensive and important task. Plenty of researches and industrial efforts have been invested on improving software testing techniques, including criteria, tools, etc. These studies can provide guidelines to select suitable test techniques for software engineers. However, in some engineering projects, business issues may be more important than technical ones, hence we need to lobby non-technical members to support our decisions. In this paper, a well-known investment model, Nelson-Siegel model, is introduced to evaluate and forecast the processes of testing with different testing criteria. Through this model, we provide a new perspective to understand short-term, medium-term, and long-term returns of investments throughout the process of testing. A preliminary experiment is conducted to investigate three testing criteria from the viewpoint of investments. The results show that statement-coverage criterion performs best in gaining long-term yields; the short-term and medium-term yields of testing depend on the scale of programs and the number of faults they contain.

cs.SE↗

arXiv2017

Generative Poisoning Attack Method Against Neural Networks

Poisoning attack is identified as a severe security threat to machine learning algorithms. In many applications, for example, deep neural network (DNN) models collect public data as the inputs to perform re-training, where the input data can be poisoned. Although poisoning attack against support vector machines (SVM) has been extensively studied before, there is still very limited knowledge about how such attack can be implemented on neural networks (NN), especially DNNs. In this work, we first examine the possibility of applying traditional gradient-based method (named as the direct gradient method) to generate poisoned data against NNs by leveraging the gradient of the target model w.r.t. the normal data. We then propose a generative method to accelerate the generation rate of the poisoned data: an auto-encoder (generator) used to generate poisoned data is updated by a reward function of the loss, and the target NN model (discriminator) receives the poisoned data to calculate the loss w.r.t. the normal data. Our experiment results show that the generative method can speed up the poisoned data generation rate by up to 239.38x compared with the direct gradient method, with slightly lower model accuracy degradation. A countermeasure is also designed to detect such poisoning attack methods by checking the loss of the target model.

cs.CR↗

arXiv2016

A New Learning Method for Inference Accuracy, Core Occupation, and Performance Co-optimization on TrueNorth Chip

IBM TrueNorth chip uses digital spikes to perform neuromorphic computing and achieves ultrahigh execution parallelism and power efficiency. However, in TrueNorth chip, low quantization resolution of the synaptic weights and spikes significantly limits the inference (e.g., classification) accuracy of the deployed neural network model. Existing workaround, i.e., averaging the results over multiple copies instantiated in spatial and temporal domains, rapidly exhausts the hardware resources and slows down the computation. In this work, we propose a novel learning method on TrueNorth platform that constrains the random variance of each computation copy and reduces the number of needed copies. Compared to the existing learning method, our method can achieve up to 68.8% reduction of the required neuro-synaptic cores or 6.5X speedup, with even slightly improved inference accuracy.

cs.NE↗

arXiv2014

Multiquark Cluster Form Factors In the Relativistic Harmonic Oscillator Model

A QCD multiquark cluster system is studied in the relativistic harmonic oscillator potential model (RHOPM), and the electromagnetic form factors of the pion, proton and deuteron in the RHOPM are predicted. The calculated theoretical results are then compared with existing experimental data, finding very good agreement between the theoretical predictions and experimental data for these three target particles. We claim that this model can be applied to study QCD hadronic properties, particularly neutron properties, and to find six-quark cluster and/or nine-quark cluster probabilities in light nuclei such as helium $^{3}He$ and tritium $^{3}H$. This is a problem of particular importance and interest in quark nuclear physics.

hep-ph↗