Year
Publisher
Journal
Institution
1
Institution Country
Publication Type
Field Of Study
Topics
Open Access
Language
Filter 1
Year
Publisher
Journal
Institution
1
Institution Country
Publication Type
Field Of Study
Topics
Open Access
Language
Filter 1
Export
Sort by: Relevance
Multi-Task Scenario Encrypted Traffic Classification and Parameter Analysis.

The widespread use of encrypted traffic poses challenges to network management and network security. Traditional machine learning-based methods for encrypted traffic classification no longer meet the demands of management and security. The application of deep learning technology in encrypted traffic classification significantly improves the accuracy of models. This study focuses primarily on encrypted traffic classification in the fields of network analysis and network security. To address the shortcomings of existing deep learning-based encrypted traffic classification methods in terms of computational memory consumption and interpretability, we introduce a Parameter-Efficient Fine-Tuning method for efficiently tuning the parameters of an encrypted traffic classification model. Experimentation is conducted on various classification scenarios, including Tor traffic service classification and malicious traffic classification, using multiple public datasets. Fair comparisons are made with state-of-the-art deep learning model architectures. The results indicate that the proposed method significantly reduces the scale of fine-tuning parameters and computational resource usage while achieving performance comparable to that of the existing best models. Furthermore, we interpret the learning mechanism of encrypted traffic representation in the pre-training model by analyzing the parameters and structure of the model. This comparison validates the hypothesis that the model exhibits hierarchical structure, clear organization, and distinct features.

Read full abstract
Open Access
Amanita peptide toxin analysis: An innovative UPLC–MS/MS method utilizing Fe3O4-based magnetic adsorbent

This study combined the structural characteristics of amanita peptide toxins to prepare a Fe3O4/P4VPBA/PPy/GO composite material as a magnetic solid-phase extraction adsorbent. The experimental results indicated that this adsorbent showed excellent selectivity for the extraction of amatoxins and phallotoxins from mushroom samples. Compared with traditional dispersive solid-phase extraction, this adsorbent enables rapid separation in the enrichment and elution recovery stages through the application of a magnetic field. Simultaneously, the study optimized the conditions for ultraperformance liquid chromatography–mass spectrometry (UPLC–MS/MS), ensuring both separation efficiency and accuracy while reducing the analysis time. The researchers established the MSPE–UPLC–MS/MS method for the determination of amanita peptide toxins in wild mushrooms. This method not only simplified and expedited the sample pretreatment process but also achieved qualitative and quantitative analyses in a short period, thus demonstrating high detection sensitivity and outstanding recovery rates. Furthermore, the developed method was successfully applied to analyze toxic mushrooms discovered in Nanjing, China. The potential for extensive utilization of this method is highlighted in the field of food analysis and detection.

Read full abstract
YOLOv8-PD:An Improved Road damage detection algorithm based on YOLOv8n model

Abstract Road damage detection is an crucial task to ensure road safety. To tackle the issues of poor performance on multi-scale cracks and high costs in road damage detection, this paper presents an improved lightweight road damage detection algorithm based on YOLOv8n, named YOLOv8-PD (Pavement Disease). Firstly, a BOT module that can extract global information of road damage images is proposed to adapt to the large-span features of crack objects. Secondly, the introduction of the Large Separable Kernel Attention (LKSA) mechanism enhances the detection accuracy of the algorithm. Then, a C2fGhost block is constructed in the neck network to strengthen the feature extraction of complex road diseases while reducing the computational load. Furthermore, we introduced Lightweight Shared Convolution Detection Head (LSCD-Head) to improve feature expressiveness and reduce the number of parameters. Finally, extensive experiments on the RDD2022 dataset yield a model with parametric and computational quantities of 2.3M and 6.1 GFLOPs, which are only 74.1% and 74.3% of the baseline, and the mAP reaches an improvement of 1.4 percentage points from the baseline. In addition, experimental results on the RoadDamage dataset show that the mAP increased by 4.2% and this algorithm has good robustness. This method can provide a reference for the automatic detection method of pavement cracks.

Read full abstract
Open Access
Research on a Framework for Chinese Argot Recognition and Interpretation by Integrating Improved MECT Models.

In underground industries, practitioners frequently employ argots to communicate discreetly and evade surveillance by investigative agencies. Proposing an innovative approach using word vectors and large language models, we aim to decipher and understand the myriad of argots in these industries, providing crucial technical support for law enforcement to detect and combat illicit activities. Specifically, positional differences in semantic space distinguish argots, and pre-trained language models' corpora are crucial for interpreting them. Expanding on these concepts, the article assesses the semantic coherence of word vectors in the semantic space based on the concept of information entropy. Simultaneously, we devised a labeled argot dataset, MNGG, and developed an argot recognition framework named CSRMECT, along with an argot interpretation framework called LLMResolve. These frameworks leverage the MECT model, the large language model, prompt engineering, and the DBSCAN clustering algorithm. Experimental results demonstrate that the CSRMECT framework outperforms the current optimal model by 10% in terms of the F1 value for argot recognition on the MNGG dataset, while the LLMResolve framework achieves a 4% higher accuracy in interpretation compared to the current optimal model.The related experiments undertaken also indicate a potential correlation between vector information entropy and model performance.

Read full abstract
Open Access
Boosting Photothermocatalytic Oxidation of Toluene Over Pt/N-TiO2: The Gear Effect of Light and Heat.

Photothermal catalysis is extremely promising for the removal of various indoor pollutants owing to its photothermal synergistic effect, while the low light utilization efficiency and unclear catalytic synergistic mechanism hinder its practical applications. Here, nitrogen atoms are introduced, and Pt nanoparticles are loaded on TiO2 to construct Pt/N-TiO2-H2, which exhibits 3.5-fold higher toluene conversion rate than the pure TiO2. Compared to both photocatalytic and thermocatalytic processes, Pt/N-TiO2-H2 exhibited remarkable performance and stability in the photothermocatalytic oxidation of toluene, achieving 98.4% conversion and 98.3% CO2 yield under a light intensity of 260 mW cm-2. Furthermore, Pt/N-TiO2-H2 demonstrated potential practical applicability in the photothermocatalytic elimination of various indoor volatile organic compounds. The synergistic effect occurs as thermocatalysis accelerates the accumulation of carboxylate species and the degradation of aldehyde species, while photocatalysis promotes the generation of aldehyde species and the consumption of carboxylate species. This ultimately enhances the photothermocatalytic process. The photothermal synergistic effect involves the specific conversion of intermediates through the interplay of light and heat, providing novel insights for the design of photothermocatalytic materials and the understanding of photothermal mechanisms.

Read full abstract
Development research of latent fingermarks based on aggregation-induced emission technique.

Fingerprints hold evidential value for individual identification; a sensitive, efficient, and convenient method for visualizing latent fingermarks (LFMs) is of great importance in the field of crime scene investigation. In this study, we proposed an aggregation-induced emission atomization technique (AIE-AT) to obtain high-quality fingermark images. Six volunteers made over 1566 fingerprint samples on 17 different objects. The quality of fingermark development was evaluated using grayscale analysis for quantitative assessment, combining the fluency of fingermark ridges and the degree of level 2 and level 3 features. Both qualitative and quantitative methods were employed to explore the effectiveness of AIE molecule C27H19N3SO in developing fingermarks, its applicability to objects, and its individual selectivity. Additionally, the stability of the AIE molecule was examined. Comparative experimental results demonstrated the high stability of the AIE molecule, making it suitable for long-term preservation. The grayscale ratio of the ridges and furrows was at least 2, with high brightness contrast, the level 2 and level 3 features were clearly observable. The AIE-AT proved to be effective for developing fingermarks on nonporous, porous, and semiporous objects. It exhibited low selectivity on suspects who leave fingermarks and showed better development effects on challenging objects, as well as efficient extraction capability for insitu fingermarks. In summary, AIE-AT can efficiently develop latent fingermarks on common objects and even challenging ones. It locates the latent fingermarks for further accurate extraction of touch exfoliated cells insitu, providing technical support for the visualization of fingermarks and the localization for extraction of touch DNA.

Read full abstract
Development of an analytical system for dried blood spots for forensic toxicology: a case study of five common drugs and poisons

Dried blood spot (DBS) technology is a simple and convenient method for collecting, transporting, and storing blood samples on filter paper, and has numerous applications in the clinical, research, and public health settings. This technique is gaining popularity in the field of forensic science because it facilitates the rapid analysis of prohibited drugs in blood samples and offers significant advantages in toxicology scenarios such as drinking-driving screening, drug abuse detection, and doping detection. However, the lack of a standardized system and the fact that its stability and reliability have not been thoroughly researched and demonstrated limit its application in judicial practice in China. DBS samples can be prepared, stored, and analyzed in various ways, all of which may significantly affect the results. In this study, we developed a method based on ultra-high performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) that focuses on the preparation, pretreatment, analysis, and storage of DBS samples. A thorough investigation was conducted to examine the optimal preparation conditions, including the blood spot matrix, drying technique, and preprocessing parameters, such as the solvent and extraction method. Moreover, the analytical conditions, such as the mobile phase system and elution gradient, were established to facilitate the quantitative detection of methamphetamine, lidocaine, ketamine, fentanyl, and diazepam in both DBS and whole-blood samples. The impact of storage conditions, such as the temperature, humidity, and sealing, on the analytical results of the DBS and whole-blood samples was also examined. The results showed a strong linear relationship for lidocaine and fentanyl within the range of 0.5-100 ng/mL. Similarly, methamphetamine, ketamine, and diazepam exhibited good linearity within the range of 2-100 ng/mL. The coefficients of determination (r2) ranged from 0.9983 to 0.9997, and the limits of detection ranged from 0.2 to 0.5 ng/mL, indicating a high degree of correlation and sensitivity. Stability tests demonstrated that the five target substances remained stable in the DBS for 60 days, with the measured contents deviating from the nominal values by 15%. Moreover, the measurement results of the DBS samples were highly similar to those of the whole-blood samples, with mean percentage differences of 4.44%, 3.50%, 7.66%, 5.10%, and 5.25% for fentanyl, diazepam, ketamine, lidocaine, and methamphetamine, respectively. Throughout the 60-day storage period, the maintenance of temperatures of -20 and 4 ℃, as well as sealing and dry storage, was not necessary. Room temperature was the most practical storage environment for the DBS samples. The results for each target showed very small concentration differences between the whole-blood and DBS samples, indicating that the DBS samples were suitable for drug and poison analysis in blood. Furthermore, the DBSs exhibited high quantitative consistency with the whole-blood samples, rendering them suitable matrices for preserving blood samples. Because DBS samples are easy to handle and store, they can realize the lightweight preservation of blood samples and provide a novel solution for the analysis and preservation of blood samples in public security practice. We recommend conducting comprehensive validations before utilizing DBS for analysis, particularly in terms of quantification, to ensure the judicial reliability of the results.

Read full abstract
Open Access