HAL-CentraleSupelec
Not a member yet
77624 research outputs found
Sort by
Target Speaker Anonymization In Multi-Speaker Recordings
No full textInternational audienceMost of the existing speaker anonymization research has focused on single-speaker audio, leading to the development of techniques and evaluation metrics optimized for such condition. This study addresses the significant challenge of speaker anonymization within multi-speaker conversational audio, specifically when only a single target speaker needs to be anonymized. This scenario is highly relevant in contexts like call centers, where customer privacy necessitates anonymizing only the customer's voice in interactions with operators. Conventional anonymization methods are often not suitable for this task. Moreover, current evaluation methodology does not allow us to accurately assess privacy protection and utility in this complex multi-speaker scenario. This work aims to bridge these gaps by exploring effective strategies for targeted speaker anonymization in conversational audio, highlighting potential problems in their development and proposing corresponding improved evaluation methodologies.</div
De-risking renewable investments: Internalizing the risks and quantifying the impact of de-risking instruments
No full textInternational audienceMobilizing sufficient investment for renewable energy is critical to achieving global climate goals, yet high financing costs – primarily driven by risk perceptions – continue to hinder the deployment of variable renewables that are capital intensive, especially in emerging economies. This study comprehensively assesses the risks affecting renewable investments, categorizing them into political, economic, transformation, resource, curtailment, and technological risks. We then map these risk categories to targeted de-risking instruments, including guarantee schemes, regulatory measures, and economic incentives. Building on an extended Capital Asset Pricing Model (CAPM), we develop and calibrate a novel cost of capital model that internalizes individual risk elements and the effect of de-risking tools across a global dataset.Our results show that, without de-risking, the cost of capital can exceed 16 % for solar PV and wind projects in high-risk countries, compared to below 6 % in advanced markets. De-risking instruments – such as political risk guarantees and tax incentives – can reduce the project cost of capital by up to 5 percentage points, leading to a 30 %–35 % reduction in the levelized cost of energy. These findings highlight the critical role of tailored de-risking strategies in accelerating clean energy transitions and offer actionable insights for policymakers and investors
Residual correction for POD-Galerkin reduced order models via conditional VAEs: Application to the bias-aware solution of forward and inverse problems
No full textInternational audiencePOD-Galerkin reduced order models (PGROMs) significantly accelerate the solution of parameterized partial differential equations (PDEs) but can suffer from accuracy limitations across the parameter space. We propose a hybrid framework that augments a PGROM with a data-driven residual corrector based on a conditional variational autoencoder (cVAE). In the offline stage, we (i) construct a standard PGROM from high-fidelity snapshots and (ii) train a cVAE on PGROM residuals defined as the differences between PGROM approximations and the corresponding highfidelity solutions. The cVAE learns a mapping from time-parameter pairs to residual fields. In the online stage, for a new parameter instance, the PGROM provides a baseline approximation while the cVAE decoder predicts the residual; adding the two yields a corrected, high-accuracy solution. The effectiveness of the proposed hybrid approach is demonstrated on three parameterized PDE problems: a heat dissipation problem with varying thermal conductivity, a pollutant transport problem with varying convection velocity, and a plane-strain elastodynamics problem with a circular inclusion whose stiffness and density are parameterized
The Third VoicePrivacy Challenge: Preserving Emotional Expressiveness and Linguistic Content in Voice Anonymization
No full textWe present results and analyses from the third VoicePrivacy Challenge held in 2024, which focuses on advancing voice anonymization technologies. The task was to develop a voice anonymization system for speech data that conceals a speaker's voice identity while preserving linguistic content and emotional state. We provide a systematic overview of the challenge framework, including detailed descriptions of the anonymization task and datasets used for both system development and evaluation. We outline the attack model and objective evaluation metrics for assessing privacy protection (concealing speaker voice identity) and utility (content and emotional state preservation). We describe six baseline anonymization systems and summarize the innovative approaches developed by challenge participants. Finally, we provide key insights and observations to guide the design of future VoicePrivacy challenges and identify promising directions for voice anonymization research.</div
The HOL-Light library of Multivariate Real analysis in Rocq
No full textInternational audienceIn this paper we present Coq-HOL-Light, a library with more than 20,000 mathematical theorems on Real analysis in Rocq. This library is a translation from the HOL-Light proof assistant. By proving the equivalence between a number of types, functions and predicates in HOL-Light and their counterpart in the Rocq standart library, we make this library directly usable in Rocq developments. We also present our indexing and retrieval tools, specially tailored for searching mathematical entities in the Coq-HOL-Light library with special facilities for the Rocq community. All tools underpinning our methodology are publicly available guarantying the reproducibility of the processes detailed in this paper. We argue that this work is relevant to researchers and practitioners who wish to harness HOL-Light's comprehensive analysis library to support the construction of Rocq developments
Error Exponents for Randomised List Decoding
No full textThis paper studies random-coding error exponents of randomised list decoding, in which the decoder randomly selects messages with probabilities proportional to the decoding metric of the codewords. The exponents (or bounds) are given for mismatched, and then particularised to matched and universal decoding metrics. Two regimes are studied: for fixed list size, we derive an ensemble-tight random-coding error exponent, and show that, for the matched metric, it does not improve the error exponent of ordinary decoding. For list sizes growing exponentially with the block-length, we provide a non-trivial lower bound to the error exponent that is tight at high rates under the matched metric
Uncertainty estimation in marker-based motion capture of the tennis serve
No full textInternational audienceMarker-based motion capture (MoCap) systems are widely used to analyse human movement. However, they are affected by measurement uncertainties, particularly marker placement errors (MPE) and soft tissue artefacts (STA). Here, we quantify the individual and combined effects of these two sources of uncertainty on joint angles and angular velocities in the case of the tennis serve. A Monte-Carlo approach was used to simulate 3000 perturbed marker trajectories for each uncertainty source and their combination. We applied a random offset for MPE, while sinusoidal perturbations were used to simulate for STA. The resulting joint kinematics were compared across all degrees of freedom. Confidence intervals (5-95 %), root mean square deviation (RMSD) and Minimal Detectable Changes (MDC) were calculated for key biomechanical variables. Results showed that STA predominantly affected angular velocities, while MPE had a greater impact on joint angles. The combined simulation consistently produced the largest variability, with mean confidence intervals ranging from 5.1° to 30.8° for joint angles and from 70.5°.s-1 to 248.5°.s-1 for joint angular velocities, and RMSD values ranging from 1.6° to 8.4° for joint angles and from 16.8°.s-1 to 68.0°.s-1 for joint angular velocities. To our knowledge, this is the first quantification of MPE and STA effects on ballistic movement kinematics. These results provide critical reference values, enabling more accurate comparisons across subjects and studies while accounting for measurement uncertainties
From Microscopic Interactions to Macroscopic Feedback: Real-Time Traffic Control via Neural Operators
No full textReal-time multi-scale traffic control requires capturing how microscopic vehicle interactions give rise to macroscopic flow dynamics and vice versa, yet existing methods either rely on computationally expensive PDE solvers or operate at the agent level without global awareness. This paper introduces a physics-informed surrogate model which aims at approximating the solution of the Aw-Rascle-Zhang equations by means of a neural operator capable of predicting macroscopic traffic evolution in real-time. The proposed approach integrates macroscopic field learning with microscopic feedback control, enabling a unified multi-scale framework in which a learned operator informs platoon behavior while respecting physical constraints. This framework embeds physics-based regularization and microscopic-macroscopic coupling within both the learning and control loops. The numerical results across dense and highdensity traffic scenarios show that the surrogate model is able to preserve the essential structure of traffic waves, maintain coherence with agent-level dynamics, and support string stable platoon responses during transient disturbances. These results demonstrate that physics-informed neural operators offer a computationally efficient alternative to classical PDE solvers for cooperative real-time intelligent transportation systems
Investigating the influence of affinity on the gaze behavior of individuals with Autism Spectrum Disorder (ASD) based on their unique autistic traits
No full textInternational audienceAutism spectrum is very wide, but autistic people share some traits such as social or language impairments. Another common characteristic is a strong passion, called an affinity, which can be anything from a movie character to a topic like history, or even a specific object. Numerous testimonies attest to the support provided by affinities for autistic individuals, offering a reassuring space in a sometimes frightening world. Clinical psychologists consider affinities as keys that can unlock language and learning for people with ASD. However, objective evidence of this role is still lacking. In this study, we used eye-tracking technology to explore the visual attention patterns of autistic individuals when presented with their affinity compared to neutral stimuli. We recruited 52 autistic participants and showed them 38 images: 10 featuring their affinity and 28 neutral ones, while recording their eye movements. Eye-tracking data provided crucial insights into how visual attention is modulated by affinities. Our results reveal significant variability in visual engagement depending on the specific autistic traits and affinities of the participants. Some showed heightened visual engagement with affinity images, while others withdrew their gaze. Some exhibited a mixed response, with both increased engagement and gaze withdrawal, and a few showed no difference between the two sets of images. These findings highlight the complex relationship between visual attention and affinities in autistic individuals, highlighting the potential of eye-tracking as a tool for understanding and leveraging these affinities in therapeutic and educational settings
