ELAA: An Ensemble-Learning-Based Adversarial Attack Targeting Image-Classification Model.

The research on image-classification-adversarial attacks is crucial in the realm of artificial intelligence (AI) security. Most of the image-classification-adversarial attack methods are for white-box settings, demanding target model gradients and network architectures, which is less practical when facing real-world cases. However, black-box adversarial attacks immune to the above limitations and reinforcement learning (RL) seem to be a feasible solution to explore an optimized evasion policy. Unfortunately, existing RL-based works perform worse than expected in the attack success rate. In light of these challenges, we propose an ensemble-learning-based adversarial attack (ELAA) targeting image-classification models which aggregate and optimize multiple reinforcement learning (RL) base learners, which further reveals the vulnerabilities of learning-based image-classification models. Experimental results show that the attack success rate for the ensemble model is about 35% higher than for a single model. The attack success rate of ELAA is 15% higher than those of the baseline methods.

The role of conformational dynamics in the activity of polymer-conjugated CalB in organic solvents.

Perennial interest in enzyme catalysis has been expanding its applicability from aqueous phases where enzymes are naturally evolved to organic solvents in which the majority of industrial chemical syntheses are carried out. Although conjugating an enzyme with a soluble polymer has been attempted to enhance enzyme activity in organic solvents, the underlying mechanism remains poorly understood in terms of the conformational dynamics and enzyme activity. Herein, we combine LF-NMR measurements and MD simulations to investigate the effects of polymer grafting on the conformational dynamics of CalB in organic solvents and the consequential impacts on the catalytic kinetics, using the lipase-catalyzed transesterification reaction as a model system. LF-NMR measurements confirm that conjugation with a soluble polymer improves the enzyme flexibility in organic solvents, leading to an increase in the catalytic efficiency of up to two orders of magnitude. MD simulations suggest that the conjugated enzyme samples a larger conformational space, compared to its native counterpart, validating the hypothesis that polymer motion enhances enzyme dynamics. These experimental and simulation results provide new insights for enhancing enzyme conformational dynamics and thereby catalytic kinetics in organic solvents.

Molecular dynamics simulations reveal how graphene oxide stabilizes and activates lipase in an anhydrous gas.

The interaction between Candida antarctica lipase B (CALB) and graphene oxide (GO) in an anhydrous gas was studied using molecular dynamics (MD) simulations augmented with a simulated annealing procedure to accelerate relaxation toward equilibrium. Three kinds of GO sheets with different oxygen contents were constructed to elucidate their effectiveness for stabilizing the active CALB conformation. It was shown that electrostatic forces are pivotal for the formation of CALB/GO complexes, and that a GO sheet with a higher oxygen content leads to stronger association with the protein. The simulation results suggest replacement of protein-binding water molecules by the GO surface, which was confirmed by thermogravimetric analysis. The CALB/GO assembly stabilizes the active enzyme conformation at elevated temperatures and, moreover, increases the protein flexibility near its active sites. The molecular details of GO interaction with CALB and the consequential effects on CALB stability and functionality are important for the development of unprecedented applications of gaseous enzymatic catalysis.

DCU-Net: a dual-channel U-shaped network for image splicing forgery detection.

The detection and location of image splicing forgery are a challenging task in the field of image forensics. It is to study whether an image contains a suspicious tampered area pasted from another image. In this paper, we propose a new image tamper location method based on dual-channel U-Net, that is, DCU-Net. The detection framework based on DCU-Net is mainly divided into three parts: encoder, feature fusion, and decoder. Firstly, high-pass filters are used to extract the residual of the tampered image and generate the residual image, which contains the edge information of the tampered area. Secondly, a dual-channel encoding network model is constructed. The input of the model is the original tampered image and the tampered residual image. Then, the deep features extracted from the dual-channel encoding network are fused for the first time, and then the tampered features with different granularity are extracted by dilation convolution, and then, the secondary fusion is carried out. Finally, the fused feature map is input into the decoder, and the predicted image is decoded layer by layer. The experimental results on Casia2.0 and Columbia datasets show that DCU-Net performs better than the latest algorithm and can accurately locate tampered areas. In addition, the attack experiments show that DCU-Net model has good robustness and can resist noise and JPEG recompression attacks.

Graphene oxide enabled long-term enzymatic transesterification in an anhydrous gas flux.

Gas-phase enzymatic catalysis has been long pursued but not yet utilized in industrial processes due to many limitations. Herein, we report a hydroxyl-rich graphene oxide (GO) aerogel that can preserve the enzymatic activity and stability in an anhydrous gas flow by providing a water-like microenvironment. Lipase immobilized in the GO aerogel exhibits a 5 to 10-fold increase in apparent activity than the lyophilized lipase powder in transesterification of geraniol and vinyl acetate in the gas phase and maintains the initial activity for more than 500 h. The solid-state circular dichroism measurement confirms that the lipase keeps its native conformation in the aerogel, and the thermogravimetric analysis shows that water molecules essential for the lipase activity can be replaced by the hydroxyl groups at the GO surface. The versatility of this method is demonstrated for two other lipases with different structures, promising unprecedented applications of enzyme-GO aerogels to gas-phase enzymatic catalysis.