Search

The application of deep neural networks in target detection has been widely adopted in various fields.However, the introduction of adversarial patch attacks, which add local perturbations to images to mislead deep neural networks, poses a significant threat to target detection systems based on vision techniques.To tackle this issue, an adversarial patch defense algorithm based on PatchTracker was proposed, leveraging the semantic differences between adversarial patches and image backgrounds.This algorithm comprised an upstream patch detector and a downstream data enhancement module.The upstream patch detector employed a YOLOV5 (you only look once-v5) model with attention mechanism to determine the locations of adversarial patches, thereby improving the detection accuracy of small-scale adversarial patches.Subsequently, the detected regions were covered with appropriate pixel values to remove the adversarial patches.This module effectively reduced the impact of adversarial examples without relying on extensive training data.The downstream data enhancement module enhanced the robustness of the target detector by modifying the model training paradigm.Finally, the image with removed patches was input into the downstream YOLOV5 target detection model, which had been enhanced through data augmentation.Cross-validation was performed on the public TT100K traffic sign dataset.Experimental results demonstrated that the proposed algorithm effectively defended against various types of generic adversarial patch attacks when compared to situations without defense measures.The algorithm improves the mean average precision (mAP) by approximately 65% when detecting adversarial patch images, effectively reducing the false negative rate of small-scale adversarial patches.Moreover, compared to existing algorithms, this approach significantly enhances the accuracy of neural networks in detecting adversarial samples.Additionally, the method exhibited excellent compatibility as it does not require modification of the downstream model structure.

Objective: We took advantage of a single-center cross-sectional study to investigate the effect of different drugs on intestinal microbiota and function in Crohn's disease. Methods: We studied the difference in fecal microbiota of Crohn's disease patients treated with mesalazine, azathioprine, and infliximab, as well as untreated patients, by metagenome and screened for differential microbiota. Further, we investigated functional differences in intestinal microbiota among the four groups. Results: Through metagenomic sequencing, we found that there was no difference between the four groups in ACE and Chao1 indices, but IFX and mesalazine improved species diversity and homogeneity compared to the untreated group, as evidenced by statistically significant differences in the Shannon index, Simpson index, and pielou_evenness. In addition, beta diversity suggested a difference between groups, but the difference was not significant. Non-parametric tests revealed differences between the four groups at the phylum level, class level, and genus level. Further analysis by LEfSe analysis revealed that the level of short-chain fatty acid-producing microbiota was increased in the treated groups, while there was no difference between the treated groups when compared to each other. Finally, the KEGG database and EggNOG database revealed that there were functional differences in intestinal microbiota among the four groups, including microbial metabolism pathway, cysteine and methionine metabolism pathway, cytoskeleton, etc. Conclusions: Mesalazine, azathioprine, and infliximab can all affect the intestinal microbiota and function in patients with Crohn's disease, and the drugs may alleviate intestinal inflammation in patients with Crohn's disease by modulating the intestinal microbiota. [ABSTRACT FROM AUTHOR]

Objective: To explore the potent effects of denatonium benzoate (DB) on Mas-related G protein-coupled receptor-B2(MrgprB2)-mediated rat mast cell degranulation. Methods: RBL-2H3 cells were treated with DB overnight, before challenged with MrgprB2 ligands substance P (SP). The release of β-hex from MrgprB2-activated RBL-2H3 was detected by substrate method. Detection of LTC4, IL-6, TNF-α and cPLA2 activity were performed by ELISA. The Ca2+ influx and the expression of RBL-2H3 MrgprB2 receptors were measured by fluorescence assay. Results: The results showed 10 µmol/L, 50 µmol/L, 80 µmol/L, 100 µmol/L DB treatments promoted β-hex and LTC4 releases from activated RBL-2H3, accompanied by increased Ca2+ mobilization and cPLA2 activation. DB treatments did not affect IL-6 and TNF-α LTC4 releases in MrgprB2-activated RBL-2H3, as well as the levels of MrgprB2 expression in mast cells. Conclusion: Taken together, DB enhanced the MrgprB2-mediated RBL-2H3 degranulation in vitro, probably by up-regulating Ca2+ mobilization in activated cells. [ABSTRACT FROM AUTHOR]

In order to make effective and rational use of waste slurry water in ready-mixed concrete enterprises, this paper puts forward the comprehensive utilization measures of slurry water through studying the basic properties of slurry water and its basic properties on the workability, strength and durability of concrete, and establishes the relationship between the density and concentration of slurry water, which can quickly determine the concentration of slurry water and timely guide the production. The results show that the slurry water has little influence on stability, but has great influence on setting time and water consumption of standard consistency. In addition, slurry water will promote the viscosity of concrete slurry increase, and affect the flow rate. In addition, when the slurry water concentration is ≤9%, it has little influence on the strength, rebound strength and carbonization of C30 concrete after reaching the standard workability through admixtures and water consumption, and the strength development law is similar to the standard. Keep the slurry water all equivalent to tap water, its solid content is also as part of tap water, concrete electric flux and monolithic slab concrete cracking are similar to the benchmark at a concentration of 9%, but the cracking time of the first concrete crack and the rapid development time of the crack are earlier than the benchmark. [ABSTRACT FROM AUTHOR]