Boosting Adversarial Training with Learnable Distribution

In recent years,various adversarial defense methods have been proposed to improve the robustness of deep neural networks.Adversarial training is one of the most potent methods to defend against adversarial attacks.However,the difference in the feature space between natural and adversarial examples hinders the accuracy and robustness of the model in adversarial training.This paper proposes a learnable distribution adversarial training method,aiming to construct the same distribution for training data utilizing the Gaussian mixture model.The distribution centroid is built to classify samples and constrain the distribution of the sample features.The natural and adversarial examples are pushed to the same distribution centroid to improve the accuracy and robustness of the model.The proposed method generates adversarial examples to close the distribution gap between the natural and adversarial examples through an attack algorithm explicitly designed for adversarial training.This algorithm gradually increases the accuracy and robustness of the model by scaling perturbation.Finally,the proposed method outputs the predicted labels and the distance between the sample and the distribution centroid.The distribution characteristics of the samples can be utilized to detect adversarial cases that can potentially evade the model defense.The effectiveness of the proposed method is demonstrated through comprehensive experiments.

辐照剂量对偏氟乙烯-六氟丙烯共聚物结构和性能的影响

采用凝胶含量、凝胶渗透色谱、核磁共振波谱、差示扫描量热分析、热重分析、拉伸性能等表征手段,研究不同辐照剂量(60 kGy, 90 kGy, 120 kGy, 160 kGy和200 kGy)对辐照交联的偏氟乙烯-六氟丙烯共聚物(P(VDF-HFP))的结构、热性能和力学性能的影响。结果表明,在60~200 kGy辐照范围内,随着辐照剂量的增大,P(VDF-HFP)中凝胶含量增加、交联度增加,而可溶解的未交联部分相对分子质量及其分布均减小;P(VDF-HFP)的熔融温度和热分解温度均降低,600℃残炭量增加;P(VDF-HFP)的力学性能在辐照剂量60 kGy时显著改善,且断裂伸长率随辐照剂量的增加而降低。本研究对P(VDF-HFP)形状记忆材料的性能提升和应用提供了有力的理论支持。

Median Filtering Forensics Scheme for Color Images Based on Quaternion Magnitude-Phase CNN

In the paper,a convolutional neural network based on quaternion transformation is proposed to detect median filtering for color images.Compared with conventional convolutional neural network,color images can be processed in a holistic manner in the proposed scheme,which makes full use of the correlation between RGB channels.And due to the use of convolutional neural network,it can effectively avoid the one-sidedness of artificial features.Experimental results have shown the scheme’s improvement over the state-of-the-art scheme on the accuracy of color image median filtering detection.

A Novel Quantum Stegonagraphy Based on Brown States

In this paper,a novel quantum steganography protocol based on Brown entangled states is proposed.The new protocol adopts the CNOT operation to achieve the transmission of secret information by the best use of the characteristics of entangled states.Comparing with the previous quantum steganography algorithms,the new protocol focuses on its anti-noise capability for the phase-flip noise,which proved its good security resisting on quantum noise.Furthermore,the covert communication of secret information in the quantum secure direct communication channel would not affect the normal information transmission process due to the new protocol’s good imperceptibility.If the number of Brown states transmitted in carrier protocol is many enough,the imperceptibility of the secret channel can be further enhanced.In aspect of capacity,the new protocol can further expand its capacity by combining with other quantum steganography protocols.Due to that the proposed protocol does not require the participation of the classic channel when it implements the transmission of secret information,any additional information leakage will not be caused for the new algorithm with good security.The detailed theoretical analysis proves that the new protocol can own good performance on imperceptibility,capacity and security.

Deep Learning for Distinguishing Computer Generated Images and Natural Images:A Survey

With the development of computer graphics,realistic computer graphics(CG)have become more and more common in our field of vision.This rendered image is invisible to the naked eye.How to effectively identify CG and natural images(NI)has been become a new issue in the field of digital forensics.In recent years,a series of deep learning network frameworks have shown great advantages in the field of images,which provides a good choice for us to solve this problem.This paper aims to track the latest developments and applications of deep learning in the field of CG and NI forensics in a timely manner.Firstly,it introduces the background of deep learning and the knowledge of convolutional neural networks.The purpose is to understand the basic model structure of deep learning applications in the image field,and then outlines the mainstream framework;secondly,it briefly introduces the application of deep learning in CG and NI forensics,and finally points out the problems of deep learning in this field and the prospects for the future.

Splicing Image and Its Localization:A Survey

With the rapid development of information technology,digital images have become an important medium for information transmission.However,manipulating images is becoming a common task with the powerful image editing tools and software,and people can tamper the images content without leaving any visible traces of splicing in order to gain personal goal.Images are easily spliced and distributed,and the situation will be a great threat to social security.The survey covers splicing image and its localization.The present status of splicing image localization approaches is discussed along with a recommendation for future research.

A Survey on Face Anti-Spoofing Algorithms

The development of artificial intelligence makes the application of face recognition more and more extensive,which also leads to the security of face recognition technology increasingly prominent.How to design a face anti-spoofing method with high accuracy,strong generalization ability and meeting practical needs is the focus of current research.This paper introduces the research progress of face anti-spoofing algorithm,and divides the existing face anti-spoofing methods into two categories:methods based on manual feature expression and methods based on deep learning.Then,the typical algorithms included in them are classified twice,and the basic ideas,advantages and disadvantages of these algorithms are analyzed.Finally,the methods of face anti-spoofing are summarized,and the existing problems and future prospects are expounded.

A Survey on Adversarial Examples in Deep Learning

Adversarial examples are hot topics in the field of security in deep learning.The feature,generation methods,attack and defense methods of the adversarial examples are focuses of the current research on adversarial examples.This article explains the key technologies and theories of adversarial examples from the concept of adversarial examples,the occurrences of the adversarial examples,the attacking methods of adversarial examples.This article lists the possible reasons for the adversarial examples.This article also analyzes several typical generation methods of adversarial examples in detail:Limited-memory BFGS(L-BFGS),Fast Gradient Sign Method(FGSM),Basic Iterative Method(BIM),Iterative Least-likely Class Method(LLC),etc.Furthermore,in the perspective of the attack methods and reasons of the adversarial examples,the main defense techniques for the adversarial examples are listed:preprocessing,regularization and adversarial training method,distillation method,etc.,which application scenarios and deficiencies of different defense measures are pointed out.This article further discusses the application of adversarial examples which currently is mainly used in adversarial evaluation and adversarial training.Finally,the overall research direction of the adversarial examples is prospected to completely solve the adversarial attack problem.There are still a lot of practical and theoretical problems that need to be solved.Finding out the characteristics of the adversarial examples,giving a mathematical description of its practical application prospects,exploring the universal method of adversarial example generation and the generation mechanism of the adversarial examples are the main research directions of the adversarial examples in the future.

Median Filtering Detection Based on Quaternion Convolutional Neural Network

Median filtering is a nonlinear signal processing technique and has an advantage in the field of image anti-forensics.Therefore,more attention has been paid to the forensics research of median filtering.In this paper,a median filtering forensics method based on quaternion convolutional neural network(QCNN)is proposed.The median filtering residuals(MFR)are used to preprocess the images.Then the output of MFR is expanded to four channels and used as the input of QCNN.In QCNN,quaternion convolution is designed that can better mix the information of different channels than traditional methods.The quaternion pooling layer is designed to evaluate the result of quaternion convolution.QCNN is proposed to features well combine the three-channel information of color image and fully extract forensics features.Experiments show that the proposed method has higher accuracy and shorter training time than the traditional convolutional neural network with the same convolution depth.

Effects of "Three Controls" Nutrient Management on Population Quality and Yield Components of Rice

[Objectives]This study was conducted to explore the control mechanism of agricultural non-point source pollution,and investigate the feasibility of promoting rice"three controls"nutrient management in Enping City.[Methods]With high-quality conventional rice as a material,such three treatments as three controls fertilization A(ZHY)and B(ZHY)and farmers conventional fertilization method FFP(ZXL)were set up,and the whole process of the late-season plot experiment was recorded.The agronomic characteristics of rice population quality and yield components during rice growth and development under the"three controls"fertilization technology were analyzed.[Results]Compared with the conventional fertilization method,the three controls A(ZHY)fertilization method improved rice yield by 27.13%,seed setting rate by 2.11%and 1 000-grain weight by 3.30%when reducing N,P and K by 27.13%,10.89%and 27.31%,respectively.In the case of three controls B(ZHY)omitting the last fertilization in the three controls fertilization method(4∶2∶3∶1),which saved the formula fertilizer by 11.25%,no difference was caused in yield,but the seed setting rate and 1 000-grain weight were still improved by 3.47%and 2.79%,respectively.Compared with the conventional fertilization method,the top first,second and third basal nodes of the three controls A(ZHY)fertilization method were shortened by 18.82%,17.06%and 20.52%,respectively,which plays an important role in combating typhoon and resisting lodging.[Conclusions]Compared with the conventional fertilization method,rice"three controls"nutrient management can improve yield and lodging resistance of rice,reduce fertilizer loss and agricultural non-point source pollution,and protect ecological environment.

An Overview of Adversarial Attacks and Defenses

In recent years,machine learning has become more and more popular,especially the continuous development of deep learning technology,which has brought great revolutions to many fields.In tasks such as image classification,natural language processing,information hiding,multimedia synthesis,and so on,the performance of deep learning has far exceeded the traditional algorithms.However,researchers found that although deep learning can train an accurate model through a large amount of data to complete various tasks,the model is vulnerable to the example which is modified artificially.This technology is called adversarial attacks,while the examples are called adversarial examples.The existence of adversarial attacks poses a great threat to the security of the neural network.Based on the brief introduction of the concept and causes of adversarial example,this paper analyzes the main ideas of adversarial attacks,studies the representative classical adversarial attack methods and the detection and defense methods.

A Survey on Adversarial Example

In recent years,deep learning has become a hotspot and core method in the field of machine learning.In the field of machine vision,deep learning has excellent performance in feature extraction and feature representation,making it widely used in directions such as self-driving cars and face recognition.Although deep learning can solve large-scale complex problems very well,the latest research shows that the deep learning network model is very vulnerable to the adversarial attack.Add a weak perturbation to the original input will lead to the wrong output of the neural network,but for the human eye,the difference between origin images and disturbed images is hardly to be notice.In this paper,we summarize the research of adversarial examples in the field of image processing.Firstly,we introduce the background and representative models of deep learning,then introduce the main methods of the generation of adversarial examples and how to defend against adversarial attack,finally,we put forward some thoughts and future prospects for adversarial examples.

A Survey on Machine Learning in Chemical Spectral Analysis

Chemical spectral analysis is contemporarily undergoing a revolution and drawing much attention of scientists owing to machine learning algorithms,in particular convolutional networks.Hence,this paper outlines the major machine learning and especially deep learning methods contributed to interpret chemical images,and overviews the current application,development and breakthrough in different spectral characterization.Brief categorization of reviewed literatures is provided for studies per application apparatus:X-Ray spectra,UV-Vis-IR spectra,Micro-scope,Raman spectra,Photoluminescence spectrum.End with the overview of existing circumstances in this research area,we provide unique insight and promising directions for the chemical imaging field to fully couple machine learning subsequently.

Transforming Health Care Through a Learning Health System Approach in the Digital Era:Chronic Kidney Disease Management in China

Introduction The last 30 years have witnessed the development of evidencebased medicine.It helps to achieve best practice by incorporating best available evidence into everyday practice.Conventionally,best evidence is generated through clinical studies such as randomized clinical trials(RCTs)and synthesized by systematic review and meta-analysis.Compared to evidence generation,fewer activities are taken to promote evidence uptake in practice.There is a long time lag between evidence and practice,and it may take up to 17 years[1].To close the gap,the concept of learning health system(LHS)was proposed in a roundtable on evidence-based medicine by the Institute of Medicine(IOM)in 2006,which provides a way to leverage data to learn knowledge and to feed it back to the frontline practice in real time[2].

Prevalence and treatment of hypertension in China： impacts of 2017 American College of Cardiology/American Heart Association High Blood Pressure Guideline

The definition, recommended treatment thresholds and targets of hypertension have been updated in the recent 2017 American College of Cardiology/American Heart Association(ACC/AHA) Guideline for the Prevention, Detection, Evaluation and Management of High Blood Pressure in Adults. However,the impacts of this guideline on Chinese population are currently unknown. In this study, we aim to provide updated data in China using criteria from the updated guideline. A multistage, stratified sampling method was used to obtain a representative sample of 50,171 adults aged 18 years and above.Hypertension was defined as an average systolic blood pressure !130 mm Hg, or an average diastolic blood pressure !80 mm Hg, or self-reported use of antihypertensive medications in past 2 weeks. The weighted prevalence of hypertension was 60.1%(95% confidence interval [CI] 59.3%–61.0%). The treatment rate and recommended treatment rate were 16.8%(95% CI 16.1%–17.6%) and 53.7%(95% CI52.6%–54.9%), respectively; the gap was more prominent among men and participants from rural areas.Hence, the adoption of the 2017 ACC/AHA guideline will lead to a substantial increase in both prevalence and number of patients needing treatment in China. The applicability of this guideline should be carefully evaluated based on evidence from Chinese population.

Socioeconomic disparities of patients with acute kidney injury:subgroup analysis of local hospitals in China from a national survey

To the Editor:Acute kidney injury(AKI)is a common public health problem worldwide,which can adversely affect patients’quality of life and even lead to death.^([1])Academic hospitals play an important role in admitting and providing treatment for patients with AKI,but limited data exist regarding the characteristics of patients in county-level local hospitals.Our research group initiated a comprehensive survey as part of the International Society of Nephrology’s"0 by 25"project(eliminating all deaths related to untreated AKI by 2025)for investigating the disease burden of AKI and its associated risk factors and prognosis through 22 province-level regions of China in 2013.