Reweighting adversarial examples during training plays an essential role in improving the robustness of neural networks,which lies in the fact that examples closer to the decision boundaries are much more vulnerable t...Reweighting adversarial examples during training plays an essential role in improving the robustness of neural networks,which lies in the fact that examples closer to the decision boundaries are much more vulnerable to being attacked and should be given larger weights.The probability margin(PM)method is a promising approach to continuously and path-independently mea-suring such closeness between the example and decision boundary.However,the performance of PM is limited due to the fact that PM fails to effectively distinguish the examples having only one misclassified category and the ones with multiple misclassified categories,where the latter is closer to multi-classification decision boundaries and is supported to be more critical in our observation.To tackle this problem,this paper proposed an improved PM criterion,called confused-label-based PM(CL-PM),to measure the closeness mentioned above and reweight adversarial examples during training.Specifi-cally,a confused label(CL)is defined as the label whose prediction probability is greater than that of the ground truth label given a specific adversarial example.Instead of considering the discrepancy between the probability of the true label and the probability of the most misclassified label as the PM method does,we evaluate the closeness by accumulating the probability differences of all the CLs and ground truth label.CL-PM shares a negative correlation with data vulnerability:data with larger/smaller CL-PM is safer/riskier and should have a smaller/larger weight.Experiments demonstrated that CL-PM is more reliable in indicating the closeness regarding multiple misclassified categories,and reweighting adversarial training based on CL-PM outperformed state-of-the-art counterparts.展开更多
The oxidative modification of myofibrillar proteins(MPs)has been identified as a crucial factor affecting meat quality during processing.We compared the effects of 4-hydroxy-2-nonenal(HNE)and lipoxygenase(LOX)-catalyz...The oxidative modification of myofibrillar proteins(MPs)has been identified as a crucial factor affecting meat quality during processing.We compared the effects of 4-hydroxy-2-nonenal(HNE)and lipoxygenase(LOX)-catalyzed linoleic acid(LA)oxidation products on the digestibility and gel properties in MPs from bighead carp.Both treatments resulted in decreased free amino acid content,reduced digestibility,and loss of amino acids in MPs.The HNE treatment enhanced gel strength by increasing hydrophobic interactions within the gel matrix.Conversely,the LOX-catalyzed LA oxidation enhanced disulfide bonds,leading to an agglomerated microstructure.Reduced myofibrillar protein(MP)solubility,primarily due to protein aggregation,was observed in the LA group but not in the HNE group.This distinction suggests that the impact on MP functionality is predominantly influenced by the protein aggregation induced by LOX-catalyzed LA oxidation rather than by HNE alone.展开更多
Films with excellent mechanical properties and antimicrobial activity have potential applications in food pres-ervation.The antimicrobial film was prepared by blending probiotic Bacillus velezensis 906 metabolites(906...Films with excellent mechanical properties and antimicrobial activity have potential applications in food pres-ervation.The antimicrobial film was prepared by blending probiotic Bacillus velezensis 906 metabolites(906),potassium sorbate(PS),and polyvinyl alcohol(PVA).The film was characterized using scanning electron mi-croscopy(SEM),Fourier transform infrared spectroscopy(FTIR),X-ray diffraction(XRD),and thermal stability analysis.The mechanical,barrier,and antimicrobial properties of the film were tested.The results of the physicochemical analysis indicated that,compared to the pure PVA film,the film’s mechanical properties decreased,while its barrier properties and hydrophobicity increased with the addition of PS and 906,respec-tively.The results of the antimicrobial tests revealed that the blend film exhibited significant antimicrobial activity against Vibrio parahaemolyticus,Psychrobacter maritimus,Oceanobacillus kimchii,Pseudomonas aeruginosa,and Listeria monocytogenes.SEM demonstrated good fusion between the components of the PVA/PS/906 film,resulting in a dense and crack-free structure.FTIR suggested that PS and 906 enhance molecular interactions within the film matrix.XRD revealed that the film was crystallized,with 906 increasing the crystallinity of the bonds between the film components.Thermal analysis indicated the thermal stability of the films was enhanced.Consequently,the PVA/PS/906 blend film shows significant potential for application in the preservation of aquatic products.展开更多
The problem of task assignment for multiple cooperating unmanned aerial vehicle(UAV) teams is considered. Multiple UAVs forming several small teams are needed to perform attack tasks on a set of predetermined ground t...The problem of task assignment for multiple cooperating unmanned aerial vehicle(UAV) teams is considered. Multiple UAVs forming several small teams are needed to perform attack tasks on a set of predetermined ground targets. A hierarchical task assignment method is presented to address the problem. It breaks the original problem down to three levels of sub-problems: target clustering, cluster allocation and target assignment. The first two sub-problems are centrally solved by using clustering algorithms and integer linear programming, respectively, and the third sub-problem is solved in a distributed and parallel manner, using a mixed integer linear programming model and an improved ant colony algorithm. The proposed hierarchical method can reduce the computational complexity of the task assignment problem considerably, especially when the number of tasks or the number of UAVs is large. Experimental results show that this method is feasible and more efficient than non-hierarchical methods.展开更多
基金supported by the National Natural Science Foundation of China (No.62072127,No.62002076,No.61906049)Natural Science Foundation of Guangdong Province (No.2023A1515011774,No.2020A1515010423)+3 种基金Project 6142111180404 supported by CNKLSTISS,Science and Technology Program of Guangzhou,China (No.202002030131)Guangdong basic and applied basic research fund joint fund Youth Fund (No.2019A1515110213)Open Fund Project of Fujian Provincial Key Laboratory of Information Processing and Intelligent Control (Minjiang University) (No.MJUKF-IPIC202101)Scientific research project for Guangzhou University (No.RP2022003).
文摘Reweighting adversarial examples during training plays an essential role in improving the robustness of neural networks,which lies in the fact that examples closer to the decision boundaries are much more vulnerable to being attacked and should be given larger weights.The probability margin(PM)method is a promising approach to continuously and path-independently mea-suring such closeness between the example and decision boundary.However,the performance of PM is limited due to the fact that PM fails to effectively distinguish the examples having only one misclassified category and the ones with multiple misclassified categories,where the latter is closer to multi-classification decision boundaries and is supported to be more critical in our observation.To tackle this problem,this paper proposed an improved PM criterion,called confused-label-based PM(CL-PM),to measure the closeness mentioned above and reweight adversarial examples during training.Specifi-cally,a confused label(CL)is defined as the label whose prediction probability is greater than that of the ground truth label given a specific adversarial example.Instead of considering the discrepancy between the probability of the true label and the probability of the most misclassified label as the PM method does,we evaluate the closeness by accumulating the probability differences of all the CLs and ground truth label.CL-PM shares a negative correlation with data vulnerability:data with larger/smaller CL-PM is safer/riskier and should have a smaller/larger weight.Experiments demonstrated that CL-PM is more reliable in indicating the closeness regarding multiple misclassified categories,and reweighting adversarial training based on CL-PM outperformed state-of-the-art counterparts.
基金supported by the National Key R&D Program of China(2022YFD2100902)the National Natural Science Foundation of China(32172133)+1 种基金the earmarked fund forCARS(CARS-45)funded by Chiang Mai University.
文摘The oxidative modification of myofibrillar proteins(MPs)has been identified as a crucial factor affecting meat quality during processing.We compared the effects of 4-hydroxy-2-nonenal(HNE)and lipoxygenase(LOX)-catalyzed linoleic acid(LA)oxidation products on the digestibility and gel properties in MPs from bighead carp.Both treatments resulted in decreased free amino acid content,reduced digestibility,and loss of amino acids in MPs.The HNE treatment enhanced gel strength by increasing hydrophobic interactions within the gel matrix.Conversely,the LOX-catalyzed LA oxidation enhanced disulfide bonds,leading to an agglomerated microstructure.Reduced myofibrillar protein(MP)solubility,primarily due to protein aggregation,was observed in the LA group but not in the HNE group.This distinction suggests that the impact on MP functionality is predominantly influenced by the protein aggregation induced by LOX-catalyzed LA oxidation rather than by HNE alone.
基金supported by grants from the Major Science and Technology Fund of Guangxi Province[AB23026035].
文摘Films with excellent mechanical properties and antimicrobial activity have potential applications in food pres-ervation.The antimicrobial film was prepared by blending probiotic Bacillus velezensis 906 metabolites(906),potassium sorbate(PS),and polyvinyl alcohol(PVA).The film was characterized using scanning electron mi-croscopy(SEM),Fourier transform infrared spectroscopy(FTIR),X-ray diffraction(XRD),and thermal stability analysis.The mechanical,barrier,and antimicrobial properties of the film were tested.The results of the physicochemical analysis indicated that,compared to the pure PVA film,the film’s mechanical properties decreased,while its barrier properties and hydrophobicity increased with the addition of PS and 906,respec-tively.The results of the antimicrobial tests revealed that the blend film exhibited significant antimicrobial activity against Vibrio parahaemolyticus,Psychrobacter maritimus,Oceanobacillus kimchii,Pseudomonas aeruginosa,and Listeria monocytogenes.SEM demonstrated good fusion between the components of the PVA/PS/906 film,resulting in a dense and crack-free structure.FTIR suggested that PS and 906 enhance molecular interactions within the film matrix.XRD revealed that the film was crystallized,with 906 increasing the crystallinity of the bonds between the film components.Thermal analysis indicated the thermal stability of the films was enhanced.Consequently,the PVA/PS/906 blend film shows significant potential for application in the preservation of aquatic products.
基金supported by the National Natural Science Foundation of China(7147205871401048)the Fundamental Research Funds for the Central Universities(2012HGZY0009)
文摘The problem of task assignment for multiple cooperating unmanned aerial vehicle(UAV) teams is considered. Multiple UAVs forming several small teams are needed to perform attack tasks on a set of predetermined ground targets. A hierarchical task assignment method is presented to address the problem. It breaks the original problem down to three levels of sub-problems: target clustering, cluster allocation and target assignment. The first two sub-problems are centrally solved by using clustering algorithms and integer linear programming, respectively, and the third sub-problem is solved in a distributed and parallel manner, using a mixed integer linear programming model and an improved ant colony algorithm. The proposed hierarchical method can reduce the computational complexity of the task assignment problem considerably, especially when the number of tasks or the number of UAVs is large. Experimental results show that this method is feasible and more efficient than non-hierarchical methods.
