From 2006 to 2008 I worked as a postdoctoral fellow at the Hebrew University of Jersualem in the department of statistics. Since 2008 I am a senior lecturer in the Shamoon College of Engineering, Israel. My main research interests include applied statistics, statistical signal processing, pattern recognition and machine learning with applications to spectroscopy and biomedical applications.
EDUCATION
Tom (Thomas) Trigano was born in Paris, France in 1978, and received an M.Sc. in engineering from the Telecom Paris Tech (France) and an M.Sc in Applied Probability from Paris VI University (France) in 2001. He recieved the Ph.D. degree in signal processing from the Telecom Paris Tech in 2005
Spectroscopic analysis reveals an opposite pattern between carnosic and rosmarinic acid concentrations in rosemary (Salvia rosmarinus) A. Mishra, A. Krief, M.M. Sahoo, A. Schachter, I. Gonda, N. Dudai, T. Trigano, I. Herrmann Spectrochimica Acta Part A Molecular and Biomolecular Spectroscopy, 2026 Carnosic acid (CA) and rosmarinic acid (RA) are the principal secondary metabolites (SMs) of rosemary ( Salvia rosmarinus Spenn. – Lamiaceae). Previous studies have examined CA and RA separately. However, their opposing CA and RA concentration trends have not been examined. This study tested the hypothesis that CA ( C CA ) and RA ( C RA ) concentrations exhibit an opposite pattern that can be rapidly and non-destructively spectrally detected using visible-near infrared spectroscopy. Hyperspectral data were acquired using three methods: field-based canopy measurements with a bare-fiber spectrometer, unmanned aerial vehicle (UAV)-borne hyperspectral imagery of the canopy, and laboratory measurements of rosemary leaf powder using a spectrometer with bifurcated-fiber. The datasets were analyzed using one-band correlations and normalized difference spectral indices. Furthermore, to identify wavelengths contributing to the opposite pattern, partial least squares regression coefficients were examined exclusively in rosemary leaf powder, as water absorption bands masked the C RA coefficients in the field-based canopy spectra. First derivatives of the standard CA and RA spectra demonstrated the opposite pattern at certain wavelengths that aligns with our spectral analysis. The opposite patterns between CA and RA are not only subjected to spectral analysis, but for nutrient–SM correlations as well, such as nitrogen correlating positively with C CA but negatively with C RA , and magnesium showing the reverse. Analyses of the hyperspectral data acquired from field grown plants, leaf powder and standard in the lab, together with nutrient–SM correlations,demonstrated the robustness of the opposite pattern. These opposite patterns demonstrated a reliable basis for non-destructive SM dynamics monitoring in rosemary.
Spectroscopic Pulse Embeddings by Contrastive Learning from Unlabeled Data for Pile-Up Analysis Congyu Lin, Xiaoying Zheng, Tom Trigano, Dima Bykhovsky, Yongxin Zhu, Li Tian Sensors, 2026 In nuclear spectroscopy, a physical phenomenon known as the pile-up effect distorts direct measurements by causing temporal overlap of detector pulses. Existing deep learning-based pile-up correction methods rely heavily on supervised training with simulated data, which often generalize poorly to real measurements due to simulation–experiment discrepancies. In this work, we propose a contrastive learning framework to learn robust and transferable representations directly from large-scale unlabeled real nuclear pulse signals. The detector output is segmented into physically complete pulse aggregations using a zero-crossing-based strategy, which serve as semantically coherent instances for representation learning. Physics-inspired data augmentations are designed to realistically model detector noise and bandwidth effects while preserving pulse area. A one-dimensional ResNet encoder is employed for efficient representation learning. The learned representations are transferred to pile-up identification and counting-rate estimation tasks. Experimental results on real nuclear radiation detection systems demonstrate that our method achieves strong performance and robustness under high counting-rate conditions, with particularly pronounced advantages in challenging peak pile-up scenarios.
Pre- and post-harvest spectral estimation of carnosic acid and rosmarinic acid in rosemary A. Mishra, A. Krief, M.M. Sahoo, A. Schachter, I. Gonda, N. Dudai, T. Trigano, I. Herrmann Computers and Electronics in Agriculture, 2026 Rosemary extracts, including carnosic and rosmarinic acids (CA and RA, respectively), are known for their antimicrobial and antioxidant capabilities. Traditional quantification methods of CA and RA (later on called selected phytochemicals) are often destructive and time-consuming. This study presents a spectral, non-destructive, and time-efficient approach for estimating selected phytochemicals in pre- and post-harvest stages. We acquired spectral data from field-grown rosemary plants, dry leaves, and powder as well as UAV-borne hyperspectral imagery. The analysis included a transformation sequence (second derivative, Yeo–Johnson, and standardization), followed by partial least squares regression (PLSR). To mimic real-life scenarios, we investigated a training–testing strategy denoted by “leave-one-day-out”, systematically excluding each day’s data from training. For CA estimation, the PLSR model achieved a coefficient of determination ( R 2 ) of 0.75 with a relative root mean square error (RRMSE) of 10.42% at the canopy level, 0.80 (RRMSE: 8.91%) for dry leaves, and 0.76 (RRMSE: 9.09%) for powder. RA estimation was challenging at the canopy level with an R 2 of 0.52 (RRMSE: 13.42%), but improved in post-harvest samples, reaching R 2 of 0.79 (RRMSE: 10.0%) for dry leaves and 0.75 (RRMSE: 9.78%) for powder. These results demonstrated the efficiency of the proposed approach. It offers a reliable alternative to traditional methods, with potential applications in agriculture and post-harvest industry.
Jensen–Tsallis divergence for supervised classification under data imbalance Antonio Squicciarini, Tom Trigano, David Luengo Machine Learning, 2025 In supervised classification problems using Deep Neural Networks, the loss function is typically based on the Kullback–Leibler divergence. However, alternative entropic divergence formulations, such as the Jensen–Shannon Divergence (JSD), have recently garnered attention for their unique properties. In this study, we delve deeper into the interpretation of the JSD and its generalized form, the Jensen–Tsallis Divergence (JTD), as alternative loss functions for supervised classification. When provided with one-hot encoded distributions for the true label probabilities, we demonstrate that these novel divergences impose an intrinsic output confidence regularization that prevents overfitting. Additionally, the q non-extensive parameter of the JTD directly influences the structure of the regularizer, offering increased flexibility in the formulation of the loss function. Through experiments conducted on artificially imbalanced versions of MNIST, Fashion-MNIST, SVHN and CIFAR-10 we showcase how JTD outperforms JSD and other traditional loss functions in terms of generalization performance, especially for highly imbalanced datasets.
Deep Learning Based Pile-Up Correction Algorithm for Spectrometric Data Under High-Count-Rate Measurements Yiwei Huang, Xiaoying Zheng, Yongxin Zhu, Tom Trigano, Dima Bykhovsky, Zikang Chen Sensors, 2025 Gamma-ray spectroscopy is essential in nuclear science, enabling the identification of radioactive materials through energy spectrum analysis. However, high count rates lead to pile-up effects, resulting in spectral distortions that hinder accurate isotope identification and activity estimation. This phenomenon highlights the need for automated and precise approaches to pile-up correction. We propose a novel deep learning (DL) framework plugging count rate information of pile-up signals with a 2D attention U-Net for energy spectrum recovery. The input to the model is an Energy–Duration matrix constructed from preprocessed pulse signals. Temporal and spatial features are jointly extracted, with count rate information embedded to enhance robustness under high count rate conditions. Training data were generated using an open-source simulator based on a public gamma spectrum database. The model’s performance was evaluated using Kullback–Leibler (KL) divergence, Mean Squared Error (MSE) Energy Resolution (ER), and Full Width at Half Maximum (FWHM). Results indicate that the proposed framework effectively predicts accurate spectra, minimizing errors even under severe pile-up effects. This work provides a robust framework for addressing pile-up effects in gamma-ray spectroscopy, presenting a practical solution for automated, high-accuracy spectrum estimation. The integration of temporal and spatial learning techniques offers promising prospects for advancing high-activity nuclear analysis applications.
GaSim: A python class to generate simulated time signals for gamma spectroscopy Zikang Chen, Dima Bykhovsky, Xiaoying Zheng, Tom Trigano, Yongxin Zhu Softwarex, 2025 The processing of nuclear pulse signals based on deep learning (DL) requires a well-labeled data set. However, the current energy spectrometers can only give users the final results, and do not allow manual labeling during the pulse signal collection process. The presented (GaSim) is a Python-based gamma pulse simulator of the raw detector electrical output signal with excellent customization capabilities. It allows customization of gamma pulse signal parameters from various aspects, making it versatile and useful for a wide range of detectors. Additionally, it provides the required labels for each generated electrical pulse at specified positions, enabling the creation of datasets for DL development.
Advanced Spectroscopy Time-Domain Signal Simulator for the Development of Machine and Deep Learning Algorithms Dima Bykhovsky, Zikang Chen, Yiwei Huang, Xiaoying Zheng, Tom Trigano IEEE Sensors Letters, 2025 Machinelearning methods, particularly deep learning (DL), have become essential for advanced signal processing. These methods often depend on annotated datasets, which can be limited or even unavailable in many cases. One area significantly affected is nuclear spectroscopy, where the lack of annotated datasets is due to the challenges of manually labeling signals recorded in the time domain. To address this issue, it is necessary to use simulators to generate annotated signals, ensuring that the generated time signals are as realistic as possible. This letter introduces a novel simulator designed to generate time-domain signals for gamma spectroscopy. Unlike traditional energy-spectrum simulators, our approach simulates raw sensor output for training advanced DL models. The simulator is analytically trackable, highly customizable, and lightweight, enabling researchers to tackle challenges, such as pile-up events and noise suppression. Case studies demonstrate its practical application in high-activity measurement scenarios.
Improving Quality of Life Through Engineering Education. A Case Study David Luengo, Albert Treytl, Stephanie Nestawal, Peter Arras, Kinga Korniejenko, Galyna Tabunshchyk, Tom Trigano 2022 IEEE European Technology and Engineering Management Summit E Tems 2022 Conference Proceedings, 2022
Grouped sparsity algorithm for multichannel intracardiac ECG synchronization European Signal Processing Conference, 2014
Cross-products LASSO David Luengo, Javier Via, Sandra Monzon, Tom Trigano, Antonio Artes-Rodriguez ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, 2013
Energy spectrum reconstruction for HPGE detectors using analytical pile-up correction ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, 2006
Pile-up correction algorithms for nuclear spectrometry Thomas Trigano, Thomas Dautremer, Eric Barat, Antoine Souloumiac ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, 2005
RECENT SCHOLAR PUBLICATIONS
Spectroscopic Pulse Embeddings by Contrastive Learning from Unlabeled Data for Pile-Up Analysis C Lin, X Zheng, T Trigano, D Bykhovsky, Y Zhu, L Tian Sensors 26 (7), 2138 , 2026 2026
Pre-and post-harvest spectral estimation of carnosic acid and rosmarinic acid in rosemary A Mishra, A Krief, MM Sahoo, A Schachter, I Gonda, N Dudai, T Trigano, ... Computers and Electronics in Agriculture 244, 111501 , 2026 2026
Doubly Stochastic Mean-Shift Clustering T Trigano, Y Sepulcre, I Lapidot arXiv preprint arXiv:2602.15393 , 2026 2026
Spectroscopic analysis reveals an opposite pattern between carnosic and rosmarinic acids concentration in rosemary (Salvia rosmarinus) A Mishra, A Krief, MM Sahoo, A Schachter, I Gonda, N Dudai, T Trigano, ... Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 127392 , 2025 2025
Stochastic mean-shift clustering I Lapidot, Y Sepulcre, T Trigano arXiv preprint arXiv:2511.09202 , 2025 2025 Citations: 3
FDP-TF: A fast two-pass trend filtering for ECG delineation T Trigano, Y Sepulcre, M Masika, A Perez, D Luengo Computers in Biology and Medicine 197, 110927 , 2025 2025 Citations: 1
Parallel implementation of spectral pileup correction and Gaussian noise suppression using CUDA heterogeneous architecture Z Chen, X Zheng, Y Zhu, T Trigano, Y Huang, Y Zhang Nuclear Instruments and Methods in Physics Research Section A: Accelerators … , 2025 2025
Jensen–Tsallis divergence for supervised classification under data imbalance A Squicciarini, T Trigano, D Luengo Machine Learning 114 (7), 162 , 2025 2025 Citations: 2
An open X-ray spectrometric dataset for deep learning-based pile-up correction C Lin, Z Chen, C Feng, S Gu, X Zheng, Y Zhu, T Trigano, D Bykhovsky International Conference on Wireless Artificial Intelligent Computing … , 2025 2025 Citations: 2
Deep Learning Based Energy Spectrum Estimation for High Counting Rate Nuclear Spectrometry Y Huang, C Lin, D Bykhovsky, T Trigano, Z Chen, X Zheng, Y Zhu IEEE Transactions on Instrumentation and Measurement , 2025 2025 Citations: 3
Deep learning based pile-up correction algorithm for spectrometric data under high-count-rate measurements Y Huang, X Zheng, Y Zhu, T Trigano, D Bykhovsky, Z Chen Sensors 25 (5), 1464 , 2025 2025 Citations: 3
Advanced spectroscopy time-domain signal simulator for the development of machine and deep learning algorithms D Bykhovsky, Z Chen, Y Huang, X Zheng, T Trigano IEEE Sensors Letters , 2025 2025 Citations: 3
GaSim: A python class to generate simulated time signals for gamma spectroscopy Z Chen, D Bykhovsky, X Zheng, T Trigano, Y Zhu SoftwareX 29, 102037 , 2025 2025 Citations: 3
Nanofilament organization in highly tough fibers based on lamin proteins Y Tzror, M Bezner, S Deri, T Trigano, K Ben-Harush Journal of the Mechanical Behavior of Biomedical Materials 160, 106748 , 2024 2024
Deep learning-based method for activity estimation from short-duration gamma spectroscopy recordings T Trigano, D Bykhovsky IEEE Transactions on Instrumentation and Measurement 73, 1-11 , 2024 2024 Citations: 7
Adaptive trend filtering for ECG denoising and delineation T Trigano, S Talala, D Luengo IEEE Journal of Biomedical and Health Informatics 27 (12), 5755-5766 , 2023 2023 Citations: 6
Parallel Pileup Correction for Nuclear Spectrometric Data on Many-Core Accelerators T Trigano Smart Computing and Communication: 7th International Conference, SmartCom … , 2023 2023
Fast algorithm for time decay estimation with applications to electrostatic ion beam traps T Trigano, Z Fradkin Measurement Science and Technology 34 (2), 025701 , 2023 2023
Intracardiac ECG pulse localization using overlapping block sparse reconstruction T Trigano, D Luengo Biomedical Signal Processing and Control 79, 103921 , 2023 2023 Citations: 3
Parallel Pileup Correction for Nuclear Spectrometric Data on Many-Core Accelerators Z Chen, X Kong, X Zheng, Y Zhu, T Trigano International Conference on Smart Computing and Communication, 258-267 , 2022 2022 Citations: 1
MOST CITED SCHOLAR PUBLICATIONS
Sparse Regression Algorithm for Activity Estimation in Spectrometry Y Sepulcre, T Trigano, Y Ritov IEEE Transactions on Signal Processing 61 (17), 4347-4359 , 2013 2013 Citations: 31
Statistical pileup correction method for HPGe detectors T Trigano, A Souloumiac, T Montagu, F Roueff, E Moulines IEEE Transactions on Signal Processing 55 (10), 4871-4881 , 2007 2007 Citations: 30
Semiparametric curve alignment and shift density estimation for biological data T Trigano, U Isserles, Y Ritov IEEE Transactions on Signal Processing 59 (5), 1970-1984 , 2011 2011 Citations: 28
Sparse spectral analysis of atrial fibrillation electrograms S Monzón, T Trigano, D Luengo, A Artes-Rodriguez 2012 IEEE International Workshop on Machine Learning for Signal Processing, 1-6 , 2012 2012 Citations: 26
Pileup correction algorithm using an iterated sparse reconstruction method T Trigano, I Gildin, Y Sepulcre IEEE Signal Processing Letters 22 (9), 1392-1395 , 2015 2015 Citations: 22
Cross-products LASSO D Luengo, J Vía, S Monzón, T Trigano, A Artés-Rodríguez 2013 IEEE International Conference on Acoustics, Speech and Signal … , 2013 2013 Citations: 21
Pile-up correction algorithms for nuclear spectrometry T Trigano, T Dautremer, E Barat, A Souloumiac Proceedings.(ICASSP'05). IEEE International Conference on Acoustics, Speech … , 2005 2005 Citations: 21
Fast digital filtering of spectrometric data for pile-up correction T Trigano, E Barat, T Dautremer, T Montagu IEEE Signal Processing Letters 22 (7), 973-977 , 2014 2014 Citations: 19
Blind analysis of atrial fibrillation electrograms: a sparsity-aware formulation D Luengo, S Monzón, T Trigano, J Vía, A Artés-Rodríguez Integrated Computer-Aided Engineering 22 (1), 71-85 , 2015 2015 Citations: 17
CoSA: An accelerated ISTA algorithm for dictionaries based on translated waveforms T Trigano, I Shevtsov, D Luengo Signal Processing 139, 131-135 , 2017 2017 Citations: 16
Traitement statistique du signal spectrométrique: étude du désempilement de spectre en énergie pour la spectrométrie Gamma T Trigano Télécom ParisTech , 2005 2005 Citations: 16
Nonparametric inference of photon energy distribution from indirect measurement É Moulines, F Roueff, A Souloumiac, T Trigano 2007 Citations: 15
An efficient method to learn overcomplete multi-scale dictionaries of ECG signals D Luengo, D Meltzer, T Trigano Applied Sciences 8 (12), 2569 , 2018 2018 Citations: 13
Sparse reconstruction algorithm for nonhomogeneous counting rate estimation T Trigano, Y Sepulcre, Y Ritov IEEE Transactions on Signal Processing 65 (2), 372-385 , 2016 2016 Citations: 13
On nonhomogeneous activity estimation in gamma spectrometry using sparse signal representation T Trigano, Y Sepulcre, M Roitman, U Aferiat 2011 IEEE Statistical Signal Processing Workshop (SSP), 649-652 , 2011 2011 Citations: 13
Grouped sparsity algorithm for multichannel intracardiac ECG synchronization T Trigano, V Kolesnikov, D Luengo, A Artés-Rodríguez 2014 22nd European Signal Processing Conference (EUSIPCO), 1537-1541 , 2014 2014 Citations: 12
Intensity estimation of spectroscopic signals with an improved sparse reconstruction algorithm T Trigano, J Cohen IEEE Signal Processing Letters 24 (5), 530-534 , 2017 2017 Citations: 10
Sparse ECG representation with a multi-scale dictionary derived from real-world signals D Luengo, D Meltzer, T Trigano 2018 41st International Conference on Telecommunications and Signal … , 2018 2018 Citations: 9
Process to isolate object of interest in image L SCHWARTZ, T Trigano, Y Bechor US Patent 10,417,772 , 2019 2019 Citations: 8
Pileup attenuation for spectroscopic signals using a sparse reconstruction M Lopatin, N Moskovitch, T Trigano, Y Sepulcre 2012 IEEE 27th Convention of Electrical and Electronics Engineers in Israel, 1-5 , 2012 2012 Citations: 8
