Abstract:Data provenance technology is capable of recording and tracking the origins of sensitive documents to prevent their leakage. Traditional network traceability methods are ineffective in tracking offline documents, and key tracing for encrypted files does not ensure reliable provenance for shared files. Existing techniques such as annotation, reverse querying, and data watermarking often require user involvement and are implemented at the application layer, resulting in inadequate security, lack of transparency and flexibility, and insufficient overall system scalability. This paper introduces an innovative script-based dynamic fingerprint provenance architecture that utilizes modifications to the Linux kernel to achieve foundational provenance, enhancing the security and transparency of document tracing. The fingerprint tracking algorithm is implemented through user scripts, improving the flexibility and effectiveness of document provenance. Additionally, the fingerprint-driven algorithm is designed to meet the demands of multi-load sharing, ensuring efficient and scalable document sharing. Upon verification, this architecture has a minimal impact on the operating system and exhibits excellent scalability. In scenarios involving single or multiple load sharing, the fingerprint-driven algorithm demonstrates transparency, real-time performance, and efficiency.

Abstract:In today’s digital age, Nginx has become the most widely used Web application server on Linux systems, holding the top position in market share. Nginx plays a critical role in ensuring user service quality, making its performance optimization crucial. Although Nginx servers are widely deployed on both X86 and ARM architectures, there is a lack of comparative analysis on performance tuning for these architectures. This study aims to fill this gap by comparing automatic system parameter tuning on Nginx across the two architectures. It identifies the performance differences between X86 and ARM in different scenarios (dynamic and static request processing). When handling dynamic requests, Nginx on the X86 architecture achieves a 99th percentile latency 515 ms lower than that on the ARM architecture, reflecting a performance improvement of 287%. Conversely, in static request processing, the ARM architecture performs better, with a 99th percentile latency 220 ms lower than that of X86, marking a 60% performance increase. These findings highlight the distinct advantages of X86 and ARM architectures in handling different loads and the significant impact of hardware architecture on Nginx performance optimization strategies. Therefore, system administrators must consider performance differences between static and dynamic requests and the unique characteristics of each architecture to achieve optimal performance.

Abstract:Effectively transferring knowledge from pre-trained models to downstream video understanding tasks is an important topic in computer vision research. Knowledge transfer becomes more challenging in open domain due to poor data conditions. Many recent multi-modal pre-training models are inspired by natural language processing and perform transfer learning by designing prompt learning. The paper leverages the comprehension ability of large language models over open domains and proposes a domain-context-assisted method for open-domain behavior recognition. This approach aligns visual representation with multi-level descriptions of human actions for robust classification, by enriching action labels with context knowledge in large language model. In the experiments of open-domain action recognition with fully supervised setting, it obtain a Top1 accuracy of 71.86% on the ARID dataset, and an mean average precision of 80.93% on the Tiny-VARIT dataset. More important, it can achieve Top1 accuracy of 48.63% in source-free video domain adaptation and 54.36% in multi-source video domain adaptation. The experimental results demonstrate the efficacy of domain context-assisted in a variety of open domain environments.

Abstract:The “Road Traffic Safety Law of the People’s Republic of China” requires that motorcycle riders and passengers must wear safety helmets as stipulated by law. Consequently, the demand for intelligent visual detection technology for helmet wearing has emerged. This paper focuses on the study of helmet wearing by riders in traffic surveillance video images, based on the YOLO object detection framework. Initially, a branch absorption module is employed to improve the residual backbone network. Subsequently, the convolutional layer feature fusion is enhanced through structural channel recombination. Finally, a designed structural fusion pruning technique is applied to further compress the model’s hyperparameters. Experimental results indicate that the algorithm boasts superior accuracy and real-time performance, with effective detection of small targets. The average precision for multi-classification reaches 88.8%, and the detection speed can achieve up to 29.5 frames per second, which essentially meets the requirements of traffic video surveillance.

Abstract:As food plays an important role in people’s daily lives, a food map showing the geographical distribution of restaurants in a city is of great social value. In the Internet Age, social media has covered every aspect of people’s lives; therefore, social media data provides a wealth of data to support automatic cartography. This work proposes an automatic generation method for urban food map driven by social media data, integrating machine learning and cartographic algorithms to realize the intelligent generation of stylized urban food map. A visualization system of urban food map has been developed, which is applied to four cities, Wuhan, Guangzhou, Chongqing, and Chengdu, for case studies. The results show the effectiveness and good visual expressiveness of this method in presenting urban cuisine for cities.

Abstract:Artificial intelligence interpretability refers to the ability of people to understand and interpret the decision-making process of machine learning models. Research in this field aims to improve the transparency of machine learning algorithms, making their decisions more trustworthy and explainable. Interpretability is crucial in artificial intelligence systems, especially in sensitive and critical decision-making domains such as healthcare, finance, and law. By providing interpretability, people can better understand the reasoning behind the model’s decisions, ensuring that they are fair, robust, and ethical. In the continuously evolving field of artificial intelligence, enhancing the interpretability of models is a key step towards achieving trustworthy and sustainable AI. This article outlines the development history of interpretable artificial intelligence and the technical characteristics of various interpretability methods, with a particular focus on interpretability in the medical field. It provides a more in-depth discussion of the limitations of current methods on medical imaging datasets and proposes possible future directions for exploration.

Abstract:Novel materials such as silicon-based, graphene, tellurium compounds, transition metal dihalogenated compounds and perovskites have unique structures and properties, and are important materials for the preparation of low-power and high-performance photodetectors. In this paper, silicon-based structures based on PN and PiN heterojunction structures are reviewed research progress of near-infrared photodetectors, as well as the latest research progress of near-infrared photodetectors based on two-dimensional materials, such as graphene, tellurium compounds, transition metal dihalogenated compounds and perovskite materials, and compare and analyze the performance parameters of related near-infrared photodetectors, which can provide ideas and references for the follow-up research of high-performance near-infrared photodetectors.