École Nationale Supérieure de Techniques Avancées

Multiscale Methods for Wave Propagation in Materials with Sign-Changing Coefficients

Abstract. From a mathematical perspective, the extraordinary properties of metamaterials are often reflected in the coefficients of the governing partial differential equations. These coefficients may fall outside the assumptions of classical theory, particularly when the effecti...

What Does It Mean for a Star-Shaped 3D Scatterer to Be Small in the Time Domain?

Abstract. In the frequency domain wave scattering problems, obstacles can be effectively replaced by point scatterers as soon as the wavelength of the incident wave exceeds significantly their diameter. The situation is less clear in the time domain, where recent works suggest th...

SuperMoTo: an integrated monitoring matrix, room correction and Ambisonics decoding plugin for audio production and immersive listening rooms

SuperMoTo is an open-source audio plugin (VST3, AU and Standalone) that puts a studio’s monitoringsignal chain into one plugin instance. From that instance it routes up to 32 inputs to 32 outputs througha configurable gain-and-EQ matrix, applies per-loudspeaker FIR correction fil...

Properties and Proximal Point Type Methods for Strongly Quasiconvex Functions in Hilbert Spaces

We study strongly quasiconvex functions defined on Hilbert spaces and taking extended real values from both theoretical and algorithmic points of view. To this end, we verify the existence of (global) minimizers and the properties of the proximal operator for such functions, prev...

Ultra-compact plasma photocathode in a hybrid wakefield accelerator

Abstract Plasma-based particle accelerators, realized by driving plasma waves with either an intense laser pulse or particle beam, offer accelerating fields orders of magnitude higher than conventional accelerators. This enables sources of highly relativistic electron beams with...

CELLServe: An SLO-Aware and Cost Efficient LLMs Serving System for Serverless Computing Environments

Large Language Models (LLMs) have enabled diverse AI applications. However, LLMs inference impose unprecedented computational and memory overhead, creating an inherent trade-off between latency Service Level Objectives (SLOs) and resource constraints. Serverless computing, with o...