Université de technologie de Troyes open archive
Not a member yet
10722 research outputs found
Sort by
Effect of forced vibrations of brake pads on the thermal performance of the disc: A numerical perspective
No full textInternational audienceIn this work we numerically investigate the thermal behaviour of automobile brake discs when the brake pads aresubjected to horizontal vibrations. To determine the influence of these vibrations on the performance of the disc, the partial differential equation (PDE) of conductive heat transfer in the disc is solved numerically using the finite element method. The outer surfaces of the disc and pad are exposed to convective and radiant heat flows. We determine the influence of the amplitude and frequency of vibrations of the plate on the isotherms’ distribution, maximum temperature, average temperature, and dissipation energy. As per the results, this study shows that adding a vibration force can reduce the heat dissipated in the disc and, therefore, improve the performance of the automobile braking system
Data Visualization and Hybrid Metaheuristics: Enhanced Particle Swarm Optimization for Healthcare
No full textInternational audienceThis article introduces an innovative approach to healthcare optimization through the integration of Particle Swarm Optimization (PSO) with other metaheuristic algorithms, leveraging the strengths of each to enhance problem-solving efficiency. By employing Exploratory Data Analysis and visual data mapping techniques, we analyze similarities and synergies among these methods to design a suitable solution method for healthcare-specific problems. The research includes a comprehensive bibliographic study, which provides a reliable data foundation for identifying effective hybrid optimization strategies. The enhanced PSO approach not only addresses complex healthcare scheduling tasks but also improves the reliability and efficiency of the proposed solutions
SIGNAL SOURCE DETECTION USING ASYMPTOTIC BEHAVIOR OF EIGENVECTORS OF RANDOM MATRICES
No full textInternational audienceThis work introduces a framework for the detection of a single source with a sensor array in the context where the noise variance and the channel between the source and the sensors are unknown at the receiver. Hypothesis test analysis is proposed. The test statistic is a function of eigenvectors and eigenvalues in the noise case. Recent results from random matrix theory are used to derive the analysis of hypothesis test, under asymptotic regime where the number of sensors and the number of observations per sensor are large but have the same order of magnitude
Scalable Universal Adversarial Watermark Defending Against Facial Forgery
No full textInternational audienceThe illegal use of facial forgery models, such as Generative Adversarial Networks (GAN) synthesized contents, has been on the rise, thereby posing great threats to personal reputation and national security. To mitigate these threats, recent studies have proposed the use of adversarial watermarks as countermeasures against GAN, effectively disrupting their outputs. However, the majority of these adversarial watermarks exhibit very limited defense ranges, providing defense against only a single GAN forgery model. Although some universal adversarial watermarks have demonstrated impressive results, they lack the defense scalability as a new-emerging forgery model appears. To address the tough issue, we propose a scalable approach even when the original forgery models are unknown. Specifically, a watermark expansion scheme, which mainly involves inheriting, defense and constraint steps, is introduced. On the one hand, the proposed method can effectively inherit the defense range of the prior well-trained adversarial watermark; on the other hand, it can defend against a new forgery model. Extensive experimental results validate the efficacy of the proposed method, exhibiting superior performance and reduced computational time compared to the state-of-the-arts
A Review of Uncertainty Management in Vehicular Platooning: Safety and Efficiency in Opportunistic Transport Environments
No full textInternational audienceIn recent years, vehicular platooning systems have gained significant attention due to their potential to enhance road safety, fuel efficiency, and traffic flow. However, these systems face several challenges, including external disturbances, communication uncertainties, cyber-attacks, and heterogeneous vehicle dynamics. This paper presents a comprehensive review of state-of-the-art control strategies designed to address these challenges in vehicular platooning systems. The focus is on various control approaches such as adaptive control, model predictive control, and robust control, which offer solutions for maintaining platoon stability despite disturbances and communication failures. We also discuss event-triggered mechanisms, which aim to reduce communication load, and cybersecurity-focused protocols that protect platooning systems from attacks. A comparison of these strategies is presented based on several criteria. The paper concludes with recommendations for future research directions to improve the robustness and real-time applicability of vehicular platooning systems in increasingly dynamic and uncertain environments
Pre-positioned inventory model for supply chain disruption mitigation
No full textInternational audienceThis paper introduces a novel pre-positioned inventory model aimed at mitigating supply chain disruptions by enhancing the resilience of supply networks characterized by multiple facilities subject to disruptions. Based on a Time-To-Recover model, we explore a single-tier supply chain framework, incorporating real-world disruption scenarios to assess the efficacy of pre-positioned inventories in disruption mitigation. A two-stage stochastic programming approach is used to formulate the problem, incorporating a special case scenario that allows for the development of a closed-form equation. This enables a detailed analysis of the impact of pre-positioned inventory on supply chain resilience, examining various scenarios to ascertain the optimal inventory levels required to mitigate disruption risks effectively. Some numerical examples are presented to illustrate the practical application of the model, offering valuable insights into the strategic positioning of inventories and the implications for supply chain design
TS-SCAN: a Density-Based Clustering Method for Trajectory Analysis and Anomaly Detection
No full textInternational audienceThe surging amount of trajectories from sensors, antennas, and radars calls for the development of modern trajectory analysis techniques. In this context, we have introduced 2 ways to compare trajectories, which include a regularisation for dynamic time warping (DTW) and a faster computation of the Fréchet distance. As an example, we present TS-SCAN, an extension of the DBSCAN algorithm, which computes trajectory clusters and their average scheme in a database
Study of an Expansion Method Based on an Image-Specific Classifier and Multi-Features for Weakly Supervised Semantic Segmentation
No full textInternational audienceIn this paper, we propose a study of an expansion method based on an image-specific classifier and multi-features for Weakly Supervised Semantic Segmentation (WSSS) with only image-level labels. Recent WSSS methods focus mainly on enhancing the pseudo masks to improve the segmentation performance by obtaining improved Class Activation Maps (CAM) or by applying post-process methods that combine expansion and refinement. Most of these methods either lack of consideration for the balance between resolution and semantics in the used features, or are carried out globally for the whole data set, without taking into account potential additional improvements based on the specific content of the image. Previously, we proposed an image-specific expansion method using multi-features to alleviate these limitations. This new study aims firstly at determining the upper performance limit of the proposed method using the ground truth masks, and secondly at analysing this performance limit in relatio n with the features chosen. Experiments show that our expansion method can achieve promising results, when used with the ground truth (upper performance) and the features that strike a balance between semantics and resolution
Detectability of Defects in the Presence of Linear Nuisance Parameters and Images Signal-Dependent Noise
No full textInternational audienceThis paper addresses two general problems of imaging systems used for visual inspection and defect detection. On the one hand, the inspected object should be carefully removed in order to detect a potential defect or an anomaly. On the other hand, one of the features of imaging systems is that the noise level depends on the image intensity and so does the detectability of defects. In addition, due to the aging of the acquisition system (LEDs, Reflector), the intensity of the illumination decreases gradually over time. The present paper addresses jointly the impact of aging imaging systems and its ensuing impact on the detectability of a defect in the presence of linear nuisance parameters and signal-dependent noise
Advancing Cough Classification: Swin Transformer vs. 2D CNN with STFT and Augmentation Techniques
No full textInternational audienceCoughing, a common symptom associated with various respiratory problems, is a crucial indicator for diagnosing and tracking respiratory diseases. Accurate identification and categorization of cough sounds, specially distinguishing between wet and dry coughs, are essential for understanding underlying health conditions. This research focuses on applying the Swin Transformer for classifying wet and dry coughs using short-time Fourier transform (STFT) representations. We conduct a comprehensive evaluation, including a performance comparison with a 2D convolutional neural network (2D CNN) model, and exploration of two distinct image augmentation methods: time mask augmentation and classical image augmentation techniques. Extensive hyperparameter tuning is performed to optimize the Swin Transformer’s performance, considering input size, patch size, embedding size, number of epochs, optimizer type, and regularization technique. Our results demonstrate the Swin Transformer’s superior accuracy, particularly when trained on classically augmented STFT images with optimized settings (320 × 320 input size, RMS optimizer, 8 × 8 patch size, and an embedding size of 128). The approach achieves remarkable testing accuracy (88.37%) and ROC AUC values (94.88%) on the challenging crowdsourced COUGHVID dataset, marking improvements of approximately 2.5% and 11% increases in testing accuracy and ROC AUC values, respectively, compared to previous studies. These findings underscore the efficacy of Swin Transformer architectures in disease detection and healthcare classification problems