ID Feature Research Articles

A robust multi-stage intrusion detection system for in-vehicle network security using hierarchical federated learning

As connected and autonomous vehicles proliferate, the Controller Area Network (CAN) bus has become the predominant communication standard for in-vehicle networks due to its speed and efficiency. However, the CAN bus lacks basic security measures such as authentication and encryption, making it highly vulnerable to cyberattacks. To ensure in-vehicle security, intrusion detection systems (IDSs) must detect seen attacks and provide a robust defense against new, unseen attacks while remaining lightweight for practical deployment. Previous work has relied solely on the CAN ID feature or has used traditional machine learning (ML) approaches with manual feature extraction. These approaches overlook other exploitable features, making it challenging to adapt to new unseen attack variants and compromising security. This paper introduces a cutting-edge, novel, lightweight, in-vehicle, IDS-leveraging, deep learning (DL) algorithm to address these limitations. The proposed IDS employs a multi-stage approach: an artificial neural network (ANN) in the first stage to detect seen attacks, and a Long Short-Term Memory (LSTM) autoencoder in the second stage to detect new, unseen attacks. To understand and analyze diverse driving behaviors, update the model with the latest attack patterns, and preserve data privacy, we propose a theoretical framework to deploy our IDS in a hierarchical federated learning (H-FL) environment. Experimental results demonstrate that our IDS achieves an F1-score exceeding 0.99 for seen attacks and exceeding 0.95 for novel attacks, with a detection rate of 99.99%. Additionally, the false alarm rate (FAR) is exceptionally low at 0.016%, minimizing false alarms. Despite using DL algorithms known for their effectiveness in identifying sophisticated and zero-day attacks, the IDS remains lightweight, ensuring its feasibility for real-world deployment. This makes our model robust against seen and unseen attacks.

Robust returns ranking prediction and portfolio optimization for M6

The M6 competition aims to address challenging problems in stock returns ranking prediction and portfolio optimization. To tackle the volatility and low signal-to-noise ratio in the stock market, our team designs the overall solution from the perspective of robustness. On the task of returns ranking prediction, we present the MultiTask Deep Neural Network with Denoising Autoencoder Enhancement (MT-DNN-DAE), which incorporates the self-supervised learning of DAE and jointly optimizes the multi-task loss. Additionally, we propose Robust Feature Selection (RFS) to identify features with a high signal-to-noise ratio for DAE’s representation learning and construct a separate branch for important ID feature to prevent its information loss. Results show our solution is capable of accurately predicting returns ranking while maintaining generalization. On the task of portfolio optimization, a Differential Evolution algorithm is also presented to optimize asset allocation and maximize returns under risk constraints, demonstrating improved performance over traditional techniques. Taken together, these methods led to a 4th place global ranking in the M6 competition.

MI[formula omitted]C: Mining intra- and inter-image context for person search

Person search aims to localize the queried person from a gallery of uncropped, realistic images. Unlike re-identification (Re-ID), person search deals with the entire scene image containing rich and diverse visual context information. However, existing works mainly focus on the person’s appearance while ignoring other essential intra- and inter-image context information. To comprehensively leverage the intra- and inter-image context, we propose a unified framework termed MI3C including the Intra-image Multi-View Context network (IMVC) and the Inter-image Group Context Ranking algorithm (IGCR). Concretely, the IMVC integrates the features from the scene, surrounding, instance, and part views collaboratively to generate the final ID feature for person search. Furthermore, the IGCR algorithm employs group matching results between query and gallery image pairs to measure the holistic image matching similarity, which is adopted as part of the sorting metric to yield a more robust ranking among the whole gallery. Extensive experiments on two popular person search benchmarks demonstrate that by mining intra- and inter-image context, our method outperforms previous state-of-the-art methods by conspicuous margins. Specifically, we achieve 96.7% mAP and 97.1% top-1 accuracy on the CUHK-SYSU dataset, 55.6% mAP, and 90.8% top-1 accuracy on the PRW dataset.

An ID-Based Robust Identification Approach Toward Multitype Noncooperative Drones

With the rapid growth of the global drone market, the security threats caused by drones have attracted widespread attention. Since legitimate and intrusive drones of multiple types may coexist in the sky, distinguishing them becomes critical for security. This paper proposes an ID-based robust individual identification approach toward multi-type non-cooperative drones. Specifically, a novel ID feature and an extraction method are proposed to obtain features from waveforms without the waveform prior knowledge. We leverage deep learning and a data augmentation method to train a robust identification model to distinguish individual drones of the same type. In order to realize the individual identification of different types of drones without increasing the model size, we design a multi-head deep learning model regarding the idea of multi-task learning. The experiment involves 18 drones of three types and their controllers. Results confirm that our approach can achieve excellent performance for multi-type individual drone identification and intrusive drone detection while being robust to noise and channel effects. It is also validated that our approach has good performance for identifying drone controllers which transmit frequency hopping signals.

An embedding and interactions learning approach for ID feature in deep recommender system

The embedding of features plays a critical role in the neural network-based recommendation model. The ID feature, which is discrete and high-dimensional sparse, contains the significant information for relevance reasoning in recommender system. However, most models use the same embedding method to address the encoding of all features. This results in inaccurate expression of ID features, as well as reduces the generalization ability of the model. Based on the assumption that learning higher-order information from the mixture of ID and other features by crossover and extraction operations provides limited benefits, or negative results, we propose an ID feature learning model (IFLM) to embed the ID tags independently and learn their interactions efficiently. IFLM separates the ID tags from the complete inputs as its exclusive input, then outputs the vectors with general data structure, thus, making it easily parallel to other deep recommendation models. Besides, IFLM generates the second-order cross feature from pairwise IDs and captures higher-order feature through the specific hidden layer, which is more efficient in learning the knowledge of ID features. We used our approach on three typical deep recommendation models. The experimental results showed that the optimized model produced better results than the corresponding baseline, in terms of prediction accuracy and convergence time.

An Embedding and Interactions Learning Approach for Id Feature in Deep Recommender System

An Embedding and Interactions Learning Approach for Id Feature in Deep Recommender System

MAPD: An improved multi-attribute pedestrian detection in a crowd

Recently, CNN (convolutional neural networks) based pedestrian detection has made significant progress, but pedestrian detection in a crowd is still a challenge. Pedestrians are close to each other in a crowd, which is difficult for discriminating individuals. To address this issue, we propose an improved Multi-attribute pedestrian detection (MAPD) method, which coptimize person intra-class compactness and inter-class discrepancy. The contributions are fourfold: (1) We analyze the effect of positive setting on the detector and adopt a better positive settings strategy to mitigate extreme class imbalance problems. (2) Inspired by Person Reid, we employ the triplet loss function to learn the advanced id feature of pedestrians. (3) We propose a novel Piecewise NMS algorithm to reduce false positive of small objects. (4) We propose a novel multi-attribute NMS algorithm based on Piecewise NMS algorithm and id information, which can adaptively distinguish predicted boxes of different pedestrians and improve the detector performance. Finally, we evaluate the MAPD detector on two benchmark datasets, including CityPersons and CrowdHuman. Results show that our approach outperforms state-of-the-art methods with a big margin.

Legal Analysis Of Copyright Issues In Youtube Content Reviewed From Law Number 28 Of 2014 On Copyright

Abstract: The purpose of this study is to know the concept of copyright protection in YouTube according to Law No. 28 of 2014 on Copyright and Knowing and analyzing the factors that cause many cases of copyright infringement in YouTube Content. The type of research used by researchers in compiling this research is a type of normative research using 2 (two) approaches, namely the Regulatory Approach and the case approach, and the Comparison Approach. The results of this study show that: First, That The Protection of Copy Right In Youtube According to Law No. 28 of 2014 on Copyright for creators whose video works are re-uploaded (reuploaded)on YouTube illegally obtained when violating the moral and or economic rights of the creator or holder of the video in the form of preventive legal protection and legal protection Repressive. Preventive legal protection is obtained by registering the creation to the Content ID feature on YouTube and also registered with the Directorate General of Intellectual Property. Repressive legal protection is obtained in the form of freedom of choice of dispute resolution path either non-litigation path or litigation path. Second, that factors that affect the number of people who commit copyright infringement or in the context of this research are violations of copyrights in Youtube content, namely: Legal Factors, Economic Pressure Factors, Cultural Factors, Education Factors, Weak Factors in law enforcement against copyright infringement perpetrators and Lack of Public Understanding of Copyright Protection.Keyword: CopyRight; Youtube; Copyright.

Entity Linking via Symmetrical Attention-Based Neural Network and Entity Structural Features

In the process of knowledge graph construction, entity linking is a pivotal step, which maps mentions in text to a knowledge base. Existing models only utilize individual information to represent their latent features and ignore the correlation between entities and their mentions. Besides, in the process of entity feature extraction, only partial latent features, i.e., context features, are leveraged to extract latent features, and the pivotal entity structural features are ignored. In this paper, we propose SA-ESF, which leverages the symmetrical Bi-LSTM neural network with the double attention mechanism to calculate the correlation between mentions and entities in two aspects: (1) entity embeddings and mention context features; (2) mention embeddings and entity description features; furthermore, the context features, structural features, and entity ID feature are integrated to represent entity embeddings jointly. Finally, we leverage (1) the similarity score between each mention and its candidate entities and (2) the prior probability to calculate the final ranking results. The experimental results on nine benchmark dataset validate the performance of SA-ESF where the average F1 score is up to 0.866.

VoIP Accessibility: A Usability Study of Voice over Internet Protocol (VoIP) Systems and a Survey of VoIP Users with Vision Loss

Introduction Accessibility of Voice over Internet Protocol (VoIP) systems was tested with a hands-on usability study and an online survey of VoIP users who are visually impaired. The survey examined the importance of common VoIP features, and both methods assessed difficulty in using those features. Methods The usability test included four paid participants who are blind and four who have low vision. Four different tasks using four different VoIP systems (two Windows-based, two iOS-based) were presented in random order. The online survey included participants with prior VoIP experience: 50 individuals who were blind and 22 who have low vision. Results Usability test participants found that receiving an incoming call was the easiest task and transferring a call was the most difficult. Those with previous iOS experience had a large advantage with the two iOS systems over those with little experience. For the online survey, most respondents (81%) had used VoIP at home for personal use and 49% had used it in their workplace. The caller ID feature was most important to participants, yet only slightly more than half found the feature easy to use; this feature was the most discrepant between importance and ease of use (p < .01). Call management was the most difficult feature for respondents. Generally, those with low vision found features to be more accessible than did those who are blind. Discussion Almost, all usability study participants reported that they could use all four systems in a real-world setting. Although some of the features were readily usable, many of the VoIP features that respondents to the online survey considered to be important were difficult for them to use. This finding indicates a serious gap between what is presently offered and what is needed for true accessibility by those with vision loss, indicating the necessity of adding accessible features to all VoIP systems. Implications for practitioners Rehabilitation staff members and teachers of young adults with vision loss should become familiar with current VoIP systems and introduce them to their clients and students in order to increase their future prospects for employment.

Disciplining the Vernacular: Fair Use, YouTube, and Remixer Agency

Disciplining the Vernacular: Fair Use, YouTube, and Remixer Agency

Medical Student Appraisal

The rapidly growing industry for mobile medical applications provides numerous smartphone resources designed for healthcare professionals. However, not all applications are equally useful in addressing the questions of early medical trainees. Three popular, free, mobile healthcare applications were evaluated along with a Google(TM) web search on both Apple(TM) and Android(TM) devices. Six medical students at a large academic hospital evaluated each application for a one-week period while on various clinical rotations. Google(TM) was the most frequently used search method and presented multimedia resources but was inefficient for obtaining clinical management information. Epocrates(TM) Pill ID feature was praised for its clinical utility. Medscape(TM) had the highest satisfaction of search and excelled through interactive educational features. Micromedex(TM) offered both FDA and off-label dosing for drugs. Google(TM) was the preferred search method for questions related to basic disease processes and multimedia resources, but was inadequate for clinical management. Caution should also be exercised when using Google(TM) in front of patients. Medscape(TM) was the most appealing application due to a broad scope of content and educational features relevant to medical trainees. Students should also be cognizant of how mobile technology may be perceived by their evaluators to avoid false impressions.

Improving contemporary algorithms for implantable cardioverter-defibrillator function

Implantable cardioverter-defibrillators (ICDs) have been shown in various clinical trials to prevent mortality from sudden cardiac death due to unstable rhythms or ventricular fibrillation. Modern ICDs use sophisticated algorithms to not only deliver therapy on the detection of a malignant rhythm but also reduce the incidence of inappropriate shocks through rhythm discrimination. Current algorithms for detection of malignant rhythms use sophisticated techniques such as real-time processing and analysis of electrograms from a transvenous lead system. The Rhythm ID feature in Boston Scientific ICDs is an example of one such algorithm used for rhythm discrimination. Rhythm ID uses the vector timing and correlation algorithm, which incorporates both timing as well as morphology information for supraventricular tachycardia discrimination. Clinical trials demonstrated high sensitivity and specificity of this feature in discriminating between ventricular tachycardia and supraventricular tachycardia (results published previously). On detection of the unknown rhythm (when the ventricular tachycardia rate detection criteria is met), the vector timing and correlation algorithm compares the unknown rhythm beat-by-beat to a stored template of normal sinus rhythm. The feature correlation coefficient computed over more than 8 points in the time-aligned signals is used for the comparison. The specific discrimination procedure of Rhythm ID depends on the mode (VR or DR) and on whether the test rhythm is an initial detected rhythm or a postshock rhythm. The normal sinus rhythm template against which the suspected rhythm is compared can be periodically updated. This article will cover some of the key aspects of the Rhythm ID feature's decision-making process and the algorithm for template update. The results of previously published clinical studies involving the algorithm's performance also will be reviewed.

Designing Features for Display-Based Systems

The purpose of this paper is to provide a detailed description of the design process used to identify the optimal implementation of the Caller ID feature for a key/pbx telephone system for small businesses. A multidisciplinary team used an assortment of design methodologies such as: literature reviews, patent searches, competitive analysis, feature simulations, focus group research, focused interviews, and usability testing to generate their final design recommendation, which has recently been awarded a patent. Specifically, this paper will focus on: 1) the competitive analysis that included a product comparison and best-in-class analysis of three commercially available Caller ID products, 2) focus group research in which participants viewed slide show demonstrations of display-based features, made ease of use and feature desirability ratings, and then expressed their views toward the features in structured focus group discussion, and 3) focused interviews with customers who had been given Caller ID display units to trial.

Modeling edges at subpixel accuracy using the local energy approach

In this paper we describe a new technique for ID and 2D edge feature extraction to subpixel accuracy using edge models and the local energy approach. A candidate edge is modeled as one of a number of parametric edge models, and the fit is refined by a least-squared error fitting technique. >