Université de technologie de Troyes open archive
Not a member yet
10722 research outputs found
Sort by
CSE: Surface Anomaly Detection with Contrastively Selected Embedding
No full textInternational audienceDetecting surface anomalies of industrial materials poses a significant challenge within a myriad of industrial manufacturing processes. In recent times, various methodologies have emerged, capitalizing on the advantages of employing a network pre-trained on natural images for the extraction of representative features. Subsequently, these features are subjected to processing through a diverse range of techniques including memory banks, normalizing flow, and knowledge distillation, which have exhibited exceptional accuracy. This paper revisits approaches based on pre-trained features by introducing a novel method centered on target-specific embedding. To capture the most representative features of the texture under consideration, we employ a variant of a contrastive training procedure that incorporates both artificially generated defective samples and anomaly-free samples during training. Exploiting the intrinsic properties of surfaces, we derived a meaningful representation from the defect-free samples during training, facilitating a straightforward yet effective calculation of anomaly scores. The experiments conducted on the MVTEC AD and TILDA datasets demonstrate the competitiveness of our approach compared to state-of-the-art methods
Precedence-based and time-indexed formulations for the flexible job shop scheduling with machine availability constraints
No full textInternational audienceReal-world scheduling problems often involve additional constraints. Machine availability constraints due to maintenance operations or breakdowns are some of the most important and difficult constraints faced by companies. In many situations, maintenance operations lead to a disruption in the production planning and require advanced optimization to maintain operations. The aim of this chapter is to introduce two mathematical models and compare their performance on a large set of instances that are made publicly available
Dynamic tailoring large-area surface plasmon polariton excitation
No full textInternational audienceAbstract We propose and demonstrate a method for dynamically changing the patterning of the surface plasmon polariton (SPP) excitation over a large area under spatially inhomogeneous polarized illumination. By illuminating a 1D gold grating with shallow rectangular grooves with a spatially structured polarization beam of near-infrared light (780 nm), we selectively excited SPPs on an extended area. The parameters used to fabricate the grating coupler, matched the wave vector of the incident light with that of the SPP to achieve an efficient coupling. The incident wave illuminating the grating is a spatially inhomogeneous polarized beam. We designed local polarization states to control the local excitation of the SPP in order to pattern large areas. For real-time local control of the polarization state of the extended incident beam, we used a setup with a spatial light modulator and quarter-wave plate
Tunability of plasmonic resonances in stratified hyperbolic metamaterials
No full textInternational audienc
Bottom-up approaches for sustainability: an empirical comparison of Frugal Innovation and Low-Tech
No full textInternational audienceHumanity faces challenges related to the ecological crisis. It is essential to reformulate the development and use of technological innovations, as well as their social meaning and environmental impact. Recently, grassroots movements and bottomup initiatives have promoted alternatives to conventional approaches to innovation. This has caught the attention of academics, professionals, and local communities and coined relevant approaches such as frugal innovation and Low-Tech. However, there are still blurred conceptual and theoretical boundaries between these two approaches. This article studies Low-Tech and Frugal Innovation for their potential to promote sustainability. Through a review of the literature, key principles and criteria were identified. Two cases were studied for an empirical comparison of their similarities and differences. This study aims to contribute to the positioning of frugal innovation and the Low-Tech in the mirror of public discourse about innovation. This article also seeks to spread and continue the debate on the potential of bottom-up approaches towards sustainability
Next-cell prediction with LSTM based on vehicle mobility for 5G mc-IoT slices
No full textInternational audienc
From synthesis to assembly: a Silicon based metasurface fabrication
No full textInternational audienceIn this study, we present an assembly technique based on the use of the capillarity force and allowing theobtaining of metasurfaces from silicon nanoparticles. Particularly, we will present the use of such techniqueto obtained complicated pattern thus paving the way to the fabrication of engineering metasurfaces
Maintenance process modelling for systems subjected to multiple failure mechanisms and maintenance actions
No full textInternational audienc
Enhanced QUBO Formulations for The Flow Shop Scheduling Problem
No full textInternational audiencen recent years, the Flow shop Scheduling Problem (FSP) has gained significant attention, prompting researchers to develop various methods, from conventional to hybrid optimization algorithms. Driven by the rapid growth of quantum computing capabilities, quantum approaches applied to optimization problems have attracted an increasing research effort. This work attempts to solve the permutation flow shop scheduling problem using quantum annealing by formulating the problem as a Quadratic Unconstrained Binary Optimization model (QUBO). Two QUBO formulations tailored to address the FSP are used. The first formulation is based on the position-based modelling while the second is based on five approximations of FSP as a Traveling Salesman Problem (TSP). The QUBO formulations are tested, using D-Wave's quantum annealers, on the well-known Taillard FSP benchmark and compared against each other. Results show that the proposed position-based QUBO reaches better solutions than the TSP-based QUBOs. This work is an attempt to highlight the increasing effectiveness of quantum annealers in addressing complex optimization problems and their limitations compared to conventional classical methods