Image classifiers that based on Deep Neural Networks(DNNs)have been proved to be easily fooled by well-designed perturbations.Previous defense methods have the limitations of requiring expensive computation or reducin...Image classifiers that based on Deep Neural Networks(DNNs)have been proved to be easily fooled by well-designed perturbations.Previous defense methods have the limitations of requiring expensive computation or reducing the accuracy of the image classifiers.In this paper,we propose a novel defense method which based on perceptual hash.Our main goal is to destroy the process of perturbations generation by comparing the similarities of images thus achieve the purpose of defense.To verify our idea,we defended against two main attack methods(a white-box attack and a black-box attack)in different DNN-based image classifiers and show that,after using our defense method,the attack-success-rate for all DNN-based image classifiers decreases significantly.More specifically,for the white-box attack,the attack-success-rate is reduced by an average of 36.3%.For the black-box attack,the average attack-success-rate of targeted attack and non-targeted attack has been reduced by 72.8%and 76.7%respectively.The proposed method is a simple and effective defense method and provides a new way to defend against adversarial samples.展开更多
The preparation of environmentally friendly oil/water separation materials remains a great challenge.Freeze-drying of wood after lignin removal yields wood aerogels,which can be used as substrates to prepare fluorine-...The preparation of environmentally friendly oil/water separation materials remains a great challenge.Freeze-drying of wood after lignin removal yields wood aerogels,which can be used as substrates to prepare fluorine-free environmentally friendly superhydrophobic materials,However,they are more suitable for absorption rather than filtration applications due to their poor strength.A study using cross-sections of pristine wood chips as substrates retains the original strength of wood,but the use of the cross-sectional of wood pieces limits their thickness,strength,and size.In this paper,a degradable fluorine-free superhydrophobic film(max.water contact angle of approximately 164.2°)with self-cleaning and abrasion resistance characteristics was prepared by a one-step method using pristine and activated walnut longitudinal section films as the substrate,with tetraethyl orthosilicate as a precursor and dodecyltriethoxysilane as a modifier.The tensile strength results show that superhydrophobic films with pristine or activated wood substrates maintained the strength of pristine wood and were 2.2 times stronger than the wood aerogel substrate.In addition,after cross-laminating the two samples,the films had the ability to separate oil and water by continuous filtration with high efficiency(98.5%)and flux(approximately 1.3×10^(3)L∙m^(‒2)∙h^(‒1)).The method has potential for the large-scale fabrication of degradable superhydrophobic filtration separation membranes.展开更多
基金The work is supported by the National Key Research Development Program of China(2016QY01W0200)the National Natural Science Foundation of China NSFC(U1636101,U1736211,U1636219).
文摘Image classifiers that based on Deep Neural Networks(DNNs)have been proved to be easily fooled by well-designed perturbations.Previous defense methods have the limitations of requiring expensive computation or reducing the accuracy of the image classifiers.In this paper,we propose a novel defense method which based on perceptual hash.Our main goal is to destroy the process of perturbations generation by comparing the similarities of images thus achieve the purpose of defense.To verify our idea,we defended against two main attack methods(a white-box attack and a black-box attack)in different DNN-based image classifiers and show that,after using our defense method,the attack-success-rate for all DNN-based image classifiers decreases significantly.More specifically,for the white-box attack,the attack-success-rate is reduced by an average of 36.3%.For the black-box attack,the average attack-success-rate of targeted attack and non-targeted attack has been reduced by 72.8%and 76.7%respectively.The proposed method is a simple and effective defense method and provides a new way to defend against adversarial samples.
基金supported by the National Natural Science Foundation of China(Grant No.51776070)the State Grid Science and Technology Program(Grant No.SGGNSW00YWJS2100024).
文摘The preparation of environmentally friendly oil/water separation materials remains a great challenge.Freeze-drying of wood after lignin removal yields wood aerogels,which can be used as substrates to prepare fluorine-free environmentally friendly superhydrophobic materials,However,they are more suitable for absorption rather than filtration applications due to their poor strength.A study using cross-sections of pristine wood chips as substrates retains the original strength of wood,but the use of the cross-sectional of wood pieces limits their thickness,strength,and size.In this paper,a degradable fluorine-free superhydrophobic film(max.water contact angle of approximately 164.2°)with self-cleaning and abrasion resistance characteristics was prepared by a one-step method using pristine and activated walnut longitudinal section films as the substrate,with tetraethyl orthosilicate as a precursor and dodecyltriethoxysilane as a modifier.The tensile strength results show that superhydrophobic films with pristine or activated wood substrates maintained the strength of pristine wood and were 2.2 times stronger than the wood aerogel substrate.In addition,after cross-laminating the two samples,the films had the ability to separate oil and water by continuous filtration with high efficiency(98.5%)and flux(approximately 1.3×10^(3)L∙m^(‒2)∙h^(‒1)).The method has potential for the large-scale fabrication of degradable superhydrophobic filtration separation membranes.