Jérôme Antoni
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Jérôme Antoni has a total of 133 co-authors. Here are the top co-authors by number of shared publications:
Liang Yu
6 shared publicationsSimon Bouley
MicrodB 4 , 69134 Écully
6 shared publications- Ge Xin
Beijing Jiaotong University
4 shared publications - Konstantinos Gryllias
KU Leuven
4 shared publications Dany Abboud
Centre technique des industries mécaniques (CETIM)
3 shared publicationsRan Wang
3 shared publications
Latest Publications
- Open Access
- 31 August 2026
- Published by Elsevier in SSRN Electronic Journal
Neuromimetic Signal Processing for Bearing Fault Diagnosis Based on Second-Order Cyclostationarity
Bearing fault condition monitoring plays an important role in the maintenance of rotating machinery. Traditional signal-processing methods have been widely used for bearing fault diagnosis; however, they often rely heavily on expert knowledge and may perform poorly in complex and noisy environments. Although neural-network-based methods have achieved remarkable success in recent years, they usually require large amounts of training data, which are not always available in real-world applications. In addition, their limited interpretability remains a major obstacle to practical deployment.To address these challenges, this paper establishes a health indicator (HI) to quantify the strength of second-order cyclostationarity (CS2) - the signature of bearing faults - and proposes a neuromimetic signal-processing (NSP) framework for bearing condition monitoring based on cyclostationary analysis. The NSP framework is designed to enhance the HI and automatically produce an optimal squared envelope spectrum (SES) for bearing fault diagnosis through a neural-network-inspired architecture. The entire process is explainable and can be implemented automatically with only one signal sample. The effectiveness of the proposed method is validated using both simulated and experimental data. The results show that the proposed approach can effectively detect weak bearing faults and can also be extended to CS2 detection in other types of signals.
- 31 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessNeuromimetic Signal Processing for Bearing Fault Diagnosis Based on Second-Order Cyclostationarity
Bearing fault condition monitoring plays an important role in the maintenance of rotating machinery. Traditional signal-processing methods have been widely used for bearing fault diagnosis; however, they often rely heavily on expert knowledge and may perform poorly in complex and noisy environments. Although neural-network-based methods have achieved remarkable success in recent years, they usually require large amounts of training data, which are not always available in real-world applications. In addition, their limited interpretability remains a major obstacle to practical deployment.To address these challenges, this paper establishes a health indicator (HI) to quantify the strength of second-order cyclostationarity (CS2) - the signature of bearing faults - and proposes a neuromimetic signal-processing (NSP) framework for bearing condition monitoring based on cyclostationary analysis. The NSP framework is designed to enhance the HI and automatically produce an optimal squared envelope spectrum (SES) for bearing fault diagnosis through a neural-network-inspired architecture. The entire process is explainable and can be implemented automatically with only one signal sample. The effectiveness of the proposed method is validated using both simulated and experimental data. The results show that the proposed approach can effectively detect weak bearing faults and can also be extended to CS2 detection in other types of signals.
- Open Access
- 2 January 2026
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 245
Optimality of the mean-square value, skewness, kurtosis, and synentropy for detection of weak additive fault signatures in (nearly) Gaussian noise
- 2 January 2026
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 245
Open AccessOptimality of the mean-square value, skewness, kurtosis, and synentropy for detection of weak additive fault signatures in (nearly) Gaussian noise
- Open Access
- 23 December 2025
- Published by EDP Sciences in Acta Acustica
- Vol. 9, 81
LVA @ INSA Lyon: half a century of research and teaching in vibration and acoustic engineering
The Laboratoire Vibrations Acoustique is a research unit of INSA Lyon, founded in the late 1960s to study vibrations and their consequences on noise emitted by machines and structures. The aim of this article is to review the historical contributions and main developments of the laboratory over the last fifty years. After examining the early years of the laboratory, the authors retrace the developments and key scientific contributions that have enabled it to gain renown at national and international levels in the field of vibration and acoustical engineering.
- 23 December 2025
- Published by EDP Sciences in Acta Acustica
- Vol. 9, 81
Open AccessLVA @ INSA Lyon: half a century of research and teaching in vibration and acoustic engineering
The Laboratoire Vibrations Acoustique is a research unit of INSA Lyon, founded in the late 1960s to study vibrations and their consequences on noise emitted by machines and structures. The aim of this article is to review the historical contributions and main developments of the laboratory over the last fifty years. After examining the early years of the laboratory, the authors retrace the developments and key scientific contributions that have enabled it to gain renown at national and international levels in the field of vibration and acoustical engineering.
- Open Access
- 26 October 2025
- Published by PHM Society in Annual Conference of the PHM Society
- Vol. 17 (1)
Bearing Spall Size Estimation Under Varying Speed Conditions
Accurate estimation of spall size in rolling element bearings is critical for effective diagnostics and prognostics in rotating machinery. Traditional methods often struggle with generalization due to noise and speed variability. This work addresses these limitations by proposing a novel approach that leverages trends in vibration measurements over time and introduces a speed-normalized condition indicator. Building on prior work, we model the bearing fault signal as a periodic pulse wave and derive a Fourier-based representation that links harmonic magnitudes to spall size. We then introduce a normalization technique using harmonic speed ratios to eliminate the influence of the system’s transfer function. Experimental validation using controlled lab data confirms the method’s ability to preserve signal extrema and improve generalizability over different speeds, offering a promising path toward scalable, real-world bearing health monitoring.
- 26 October 2025
- Published by PHM Society in Annual Conference of the PHM Society
- Vol. 17 (1)
Open AccessBearing Spall Size Estimation Under Varying Speed Conditions
Accurate estimation of spall size in rolling element bearings is critical for effective diagnostics and prognostics in rotating machinery. Traditional methods often struggle with generalization due to noise and speed variability. This work addresses these limitations by proposing a novel approach that leverages trends in vibration measurements over time and introduces a speed-normalized condition indicator. Building on prior work, we model the bearing fault signal as a periodic pulse wave and derive a Fourier-based representation that links harmonic magnitudes to spall size. We then introduce a normalization technique using harmonic speed ratios to eliminate the influence of the system’s transfer function. Experimental validation using controlled lab data confirms the method’s ability to preserve signal extrema and improve generalizability over different speeds, offering a promising path toward scalable, real-world bearing health monitoring.
- 23 August 2025
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 238
Cyclostationarity of the noise radiated by a rotating acoustic source in a duct: Proof of concept
- 23 August 2025
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 238
Cyclostationarity of the noise radiated by a rotating acoustic source in a duct: Proof of concept
- 7 August 2025
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 238
Spectrum forming and its high-performance implementations for the blade tip timing signal processing
- 7 August 2025
- Published by Elsevier in Mechanical Systems and Signal Processing
- Vol. 238
Spectrum forming and its high-performance implementations for the blade tip timing signal processing
- Open Access
- 1 July 2025
- Published by Acoustical Society of America (ASA) in The Journal of the Acoustical Society of America
- Vol. 158 (1) , 759-775
Enhancing blind separation of incoherent sources in acoustic imaging with deep learning
Separating multiple incoherent sound sources presents a significant challenge in acoustic imaging. Existing methods, such as principal component analysis (PCA) applied to the cross-spectral matrix, yield “virtual” sources based on decorrelation. However, this approach often fails due to the existence of multiple solutions in the general case. State-of-the-art methods to address this issue involve computing an additional rotation matrix to enforce criteria such as least spatial entropy, or spatial orthogonality, a process that, while effective, significantly increases computational complexity. This work introduces a hybrid approach combining PCA and deep learning for separating source maps from virtual sources. By simulating sound sources in random quantities and locations, a neural network tailored to this task is trained. The permutation problem between PCA-derived virtual sources and pre-simulated labels, as well as the estimation of the number of sources, is addressed by framing source separation as a set prediction problem utilizing the Hungarian loss. This method is array geometry agnostic and frequency agnostic, allowing for robust performance across diverse array configurations and a wide frequency range. Trained solely on simulated data, the model demonstrates effective source separation on real-world datasets, highlighting the potential of integrating deep learning with existing methods.
- 1 July 2025
- Published by Acoustical Society of America (ASA) in The Journal of the Acoustical Society of America
- Vol. 158 (1) , 759-775
Open AccessEnhancing blind separation of incoherent sources in acoustic imaging with deep learning
Separating multiple incoherent sound sources presents a significant challenge in acoustic imaging. Existing methods, such as principal component analysis (PCA) applied to the cross-spectral matrix, yield “virtual” sources based on decorrelation. However, this approach often fails due to the existence of multiple solutions in the general case. State-of-the-art methods to address this issue involve computing an additional rotation matrix to enforce criteria such as least spatial entropy, or spatial orthogonality, a process that, while effective, significantly increases computational complexity. This work introduces a hybrid approach combining PCA and deep learning for separating source maps from virtual sources. By simulating sound sources in random quantities and locations, a neural network tailored to this task is trained. The permutation problem between PCA-derived virtual sources and pre-simulated labels, as well as the estimation of the number of sources, is addressed by framing source separation as a set prediction problem utilizing the Hungarian loss. This method is array geometry agnostic and frequency agnostic, allowing for robust performance across diverse array configurations and a wide frequency range. Trained solely on simulated data, the model demonstrates effective source separation on real-world datasets, highlighting the potential of integrating deep learning with existing methods.
- 1 February 2025
- Published by Acoustical Society of America (ASA) in The Journal of the Acoustical Society of America
- Vol. 157 (2) , 1428-1447
An intrusive Gibbs sampling method for implementing the nonsynchronous measurements of microphone array
The nonsynchronous measurements of microphone array are a powerful method for achieving large array or high microphone density by scanning the object from a sequential movement of a prototype array. It has attracted great interest recently because it is beyond the fundamental limitation of working frequency that is determined by the aperture and microphone density of an array. A crucial problem of nonsynchronous measurements is to recover the missing phase relation information between consecutive positions. The problem in traditional solution is generally attributed to the matrix completion problem of a block diagonal spectral matrix. In this paper, the issue has been investigated as a problem of solving a system of equations in the Bayesian formalism. The intrusive Gibbs sampling method is proposed to reconstruct the source in the equations. In the numerical simulations, convergence diagnosis of the Markov chain is illustrated through three approaches. Acoustical source reconstruction error is also discussed with respect to the frequency range, signal-to-noise ratio, measurement distances, and shift distance of sequential movement. The proposed method is exhibited to return similar results with the expectation maximization algorithm of nonsynchronous measurements. The effectiveness of the proposed method is also validated by experiments in a semi-anechoic chamber.
- 1 February 2025
- Published by Acoustical Society of America (ASA) in The Journal of the Acoustical Society of America
- Vol. 157 (2) , 1428-1447
An intrusive Gibbs sampling method for implementing the nonsynchronous measurements of microphone array
The nonsynchronous measurements of microphone array are a powerful method for achieving large array or high microphone density by scanning the object from a sequential movement of a prototype array. It has attracted great interest recently because it is beyond the fundamental limitation of working frequency that is determined by the aperture and microphone density of an array. A crucial problem of nonsynchronous measurements is to recover the missing phase relation information between consecutive positions. The problem in traditional solution is generally attributed to the matrix completion problem of a block diagonal spectral matrix. In this paper, the issue has been investigated as a problem of solving a system of equations in the Bayesian formalism. The intrusive Gibbs sampling method is proposed to reconstruct the source in the equations. In the numerical simulations, convergence diagnosis of the Markov chain is illustrated through three approaches. Acoustical source reconstruction error is also discussed with respect to the frequency range, signal-to-noise ratio, measurement distances, and shift distance of sequential movement. The proposed method is exhibited to return similar results with the expectation maximization algorithm of nonsynchronous measurements. The effectiveness of the proposed method is also validated by experiments in a semi-anechoic chamber.