Publications associées à Structured Sparsity Models for Reverberant Speech Separation

Statistical Inference for Inverse Problems: From Sparsity-Based Methods to Neural Networks

In inverse problems, the task is to reconstruct an unknown signal from its possibly noise-corrupted measurements. Penalized-likelihood-based estimation and Bayesian estimation are two powerful statistical paradigms for the resolution of such problems. They ...

EPFL2024

Aggregating Spatial and Photometric Context for Photometric Stereo

David Honzátko

Photometric stereo, a computer vision technique for estimating the 3D shape of objects through images captured under varying illumination conditions, has been a topic of research for nearly four decades. In its general formulation, photometric stereo is an ...

EPFL2024

Equivalent black carbon concentration measured with an aethalometer AE33 during the Arctic Ocean 2018 expedition

Benjamin Jérémy Laurent Heutte, Julia Schmale, Andrea Baccarini

Equivalent black carbon (eBC) mass concentration was measured with a commercial aethalometer (model AE33, Magee Scientific, Berkeley, USA). Measurements were performed onboard of the Swedish icebreaker (I/B) Oden from August to September 2018 as part of th ...

EPFL Infoscience2024

Sparse Autoencoders for Speech Modeling and Recognition

Selen Hande Kabil

Speech recognition-based applications upon the advancements in artificial intelligence play an essential role to transform most aspects of modern life. However, speech recognition in real-life conditions (e.g., in the presence of overlapping speech, varyin ...

EPFL2023

Aerosol optical absorption coefficients at seven wavelengths in 10 min resolution measured in the Swiss container during MOSAiC 2019/2020

Benjamin Jérémy Laurent Heutte, Julia Schmale, Ivo Fabio Beck, Lubna Dada

This dataset contains aerosol optical absorption coefficients at seven different wavelengths (babs(λ)), averaged to 10 min time resolution, measured during the year-long MOSAiC expedition from October 2019 to September 2020. The measurements were performed ...

Zenodo2023

Towards learning-based denoising of light fields

Touradj Ebrahimi, Michela Testolina, Tomás Soares De Carvalho Feith

In recent years, new emerging immersive imaging modalities, e.g. light fields, have been receiving growing attention, becoming increasingly widespread over the years. Light fields are often captured through multi-camera arrays or plenoptic cameras, with th ...

2023

The inductive bias of deep learning: Connecting weights and functions

Guillermo Ortiz Jimenez

Years of a fierce competition have naturally selected the fittest deep learning algorithms. Yet, although these models work well in practice, we still lack a proper characterization of why they do so. This poses serious questions about the robustness, trus ...

EPFL2023

Massively parallel nodal discontinous Galerkin finite element method simulator for room acoustics

Jan Sickmann Hesthaven

We present a massively parallel and scalable nodal discontinuous Galerkin finite element method (DGFEM) solver for the time-domain linearized acoustic wave equations. The solver is implemented using the libParanumal finite element framework with extensions ...

Sage Publications Ltd2023

Novel Methods For Detection And Analysis Of Atypical Aspects In Speech

Julian David Fritsch

Atypical aspects in speech concern speech that deviates from what is commonly considered normal or healthy. In this thesis, we propose novel methods for detection and analysis of these aspects, e.g. to monitor the temporary state of a speaker, diseases tha ...

EPFL2023

A model of non-Maxwellian electron distribution function for the analysis of ECE data in JET discharges

Mikhail Maslov, Javier García Hernández, Samuele Mazzi

Recent experiments performed in JET at high level of plasma heating, in preparation of, and during the DT campaign have shown significant discrepancies between electron temperature measurements by Thomson Scattering (TS) and Electron Cyclotron Emission (EC ...

E D P SCIENCES2023

PID-like active impedance control for electroacoustic resonators to design tunable single-degree-of-freedom sound absorbers

Hervé Lissek, Maxime Volery, Xinxin Guo

Sound absorption at low frequencies still remains a challenge in both scientific research and engineering practice. Natural porous materials are ineffective in this frequency range, as well as acoustic resonators which present too narrow bandwidth of absor ...

2022

Corona discharge actuator as an active sound absorber under normal and oblique incidence

Thomas Humbert, Hervé Lissek, Stanislav Sergeev

In the majority of active sound absorbing systems, a conventional electrodynamic loudspeaker is used as a controlled source. However, particular situations may require an actuator that is more resistant to harsh environments, adjustable in shape, and light ...

2022

Variational Methods For Continuous-Domain Inverse Problems: the Quest for the Sparsest Solution

Thomas Jean Debarre

The goal of this thesis is to study continuous-domain inverse problems for the reconstruction of sparse signals and to develop efficient algorithms to solve such problems computationally. The task is to recover a signal of interest as a continuous function ...

EPFL2022

Light absorption by brown carbon over the South-East Atlantic Ocean

Athanasios Nenes, Lu Zhang

Biomass burning emissions often contain brown carbon (BrC), which represents a large family of light-absorbing organics that are chemically complex, thus making it difficult to estimate their absorption of incoming solar radiation, resulting in large uncer ...

Copernicus GmbH2022

Automatic pathological speech assessment

Parvaneh Janbakhshi

Many pathologies cause impairments in the speech production mechanism resulting in reduced speech intelligibility and communicative ability. To assist the clinical diagnosis, treatment and management of speech disorders, automatic pathological speech asses ...

EPFL2022

Plasma-based Electroacoustic Actuator for Broadband Sound Absorption

Stanislav Sergeev

Environmental noise, mostly related to human activities, has an immense impact on public health. The development of noise reduction technologies is paramount in addressing this problem. Because of practical and economic reasons, a compact, broadband, light ...

EPFL2022

From Undercomplete to Sparse Overcomplete Autoencoders to Improve LF-MMI Speech Recognition

Hervé Bourlard, Selen Hande Kabil

Starting from a strong Lattice-Free Maximum Mutual Information (LF-MMI) baseline system, we explore different autoencoder configurations to enhance Mel-Frequency Cepstral Coefficients (MFCC) features. Autoencoders are expected to generate new MFCC features ...

ISCA-INT SPEECH COMMUNICATION ASSOC2022

A Structured Dictionary Perspective on Implicit Neural Representations

Pascal Frossard, Beril Besbinar, Guillermo Ortiz Jimenez, Gizem Yüce

Implicit neural representations (INRs) have recently emerged as a promising alternative to classical discretized representations of signals. Nevertheless, despite their practical success, we still do not understand how INRs represent signals. We propose a ...

IEEE COMPUTER SOC2022

On Breathing Pattern Information in Synthetic Speech

Mathew Magimai Doss, Zohreh Mostaani

The respiratory system is an integral part of human speech production. As a consequence, there is a close relation between respiration and speech signal, and the produced speech signal carries breathing pattern related information. Speech can also be gener ...

ISCA-INT SPEECH COMMUNICATION ASSOC2022

Unsupervised Voice Activity Detection by Modeling Source and System Information using Zero Frequency Filtering

Mathew Magimai Doss, Eklavya Sarkar

Voice activity detection (VAD) is an important pre-processing step for speech technology applications. The task consists of deriving segment boundaries of audio signals which contain voicing information. In recent years, it has been shown that voice source ...

ISCA2022