Curriculum · LISTIC · Université Savoie Mont-Blanc

Ammar Mian

Associate professor (Maître de conférences) at LISTIC, Polytech Annecy-Chambéry, Université Savoie Mont-Blanc, working on statistical signal processing and Riemannian-geometry methods for robust and remote-sensing problems.

This yearOn a CNRS research delegation for 2026–2027, with the GAIA team at GIPSA-lab, Grenoble.

ammar.mian@univ-smb.fr·HAL·GitHub

1Career

Positions

  • 2026—2027
    Délégation CNRSGAIA team, GIPSA-lab, Grenoble — research delegation for the year
  • 2020—
    Maître de conférencesLISTIC, Polytech Annecy-Chambéry, Université Savoie Mont-Blanc
  • 2019—2020
    Post-doctorantAalto University, Finland — Riemannian geometry for statistical learning, with Esa Ollila
  • 2016—2019
    DoctorantSONDRA, Université Paris-Saclay — SAR image time-series analysis

Education

  • 2019
    DoctoratContributions à l'analyse de séries temporelles d'images SAR — SONDRA, Université Paris-Saclay
  • 2016
    Diplôme d'ingénieurGrenoble INP – Phelma, Traitement du Signal

2Publications

From the HAL open archive, as of 4 November 2026. The library in the tower has the same list with search, filters and BibTeX export.

20271 record

  1. On batch normalization for SPDnet Matthieu Gallet, Ammar Mian, Florent Bouchard, Guillaume Ginolhac Signal Processing 250, art. 110793Journal HALPDFDOI

20266 records

  1. Generalized Nonnegative Structured Kruskal Tensor Regression Xinjue Wang, Esa Ollila, Sergiy Vorobyov, Ammar Mian Signal Processing 240, art. 110338Journal HALPDFDOI
  2. pyRiemann Alexandre Barachant, Quentin Barthélemy, Gabriel Wagner Vom Berg, Alexandre Gramfort, Sylvain Chevallier, Jean-Rémi King, Pedro Rodrigues, Grégoire Cattan, Emanuele Olivetti, Vladislav Goncharenko, Anton Andreev, Tim Näher, Apolline Mellot, Ammar Mian, Steven Mortier, Marie-Constance Corsi, Jona Sassenhagen, Thibault de Surrel, Richard Höchenberger, Pierre Clisson, Igor Carrara, Frida Heskebeck SOFTWARE HALDOI
  3. Robust Multiping Forward-Looking Sonar Coherent Detection Olivier Lerda, Guillaume Ginolhac, Ammar Mian, Jean-Philippe Ovarlez, Guillaume Matte, Didier Charlot IEEE Journal of Oceanic Engineering 51(3), pp. 2258–2273Journal HALPDFDOI
  4. ARMAGNAC: A New Parametric Batch Normalization Layer for SPDNet Architecture Matthieu Gallet, Florent Bouchard, Ammar Mian, Guillaume Ginolhac EUSIPCOConference HALPDF
  5. SPD Learn: A Geometric Deep Learning Python Library for Neural Decoding Through Trivialization Bruno Aristimunha, Ce Ju, Antoine Collas, Florent Bouchard, Ammar Mian, Bertrand Thirion, Sylvain Chevallier, Reinmar Kobler Preprint HAL
  6. FedSPDnet: Geometry-Aware Federated Deep Learning with SPDnet Thibault Pautrel, Florent Bouchard, Ammar Mian, Guillaume Ginolhac EUSIPCOConference HALPDF

20253 records

  1. Classification of Buried Objects From Ground Penetrating Radar Images by Using Second-Order Deep Learning Models Douba Jafuno, Ammar Mian, Guillaume Ginolhac, Nickolas Stelzenmuller IEEE JSTARS 18, pp. 3185–3197Journal HALPDFDOI
  2. Détection et Localisation des Fissures Volcaniques dans les Interférogrammes par IA Thomas Petellat, Christophe Lin Kwong Chon, Yajing Yan, Ammar Mian Rencontre Scientifique VolcanologiqueConference HAL
  3. EFFECTS OF CLASS DISTRIBUTION SHIFT IN MULTI-LABEL CLASSIFICATION SPLITS: A STUDY CASE ON BIGEARTHNET Matthieu Verlynde, Ammar Mian, Yajing Yan IGARSSConference HALPDF

20245 records

  1. RENYI DIVERGENCES LEARNING FOR EXPLAINABLE CLASSIFICATION OF SAR IMAGE PAIRS Matthieu Gallet, Ammar Mian, Abdourrahmane Atto ICASSP, pp. 7445–7449Conference HALPDFDOI
  2. Random matrix theory improved Fréchet mean of symmetric positive definite matrices Florent Bouchard, Ammar Mian, Malik Tiomoko, Guillaume Ginolhac, Frédéric Pascal ICMLConference HALPDF
  3. Online change detection in SAR time-series with Kronecker product structured scaled Gaussian models Ammar Mian, Guillaume Ginolhac, Florent Bouchard, Arnaud Breloy Signal Processing 224, art. 109589Journal HALPDFDOI
  4. Buried Object Classification from GPR Data by using Second Order Deep Learning Models Douba Jafuno, Ammar Mian, Guillaume Ginolhac, Nickolas Stelzenmuller IGARSSConference HALPDF
  5. Robust Detection for Mills Cross Sonar Olivier Lerda, Ammar Mian, Guillaume Ginolhac, Jean-Philippe Ovarlez, Didier Charlot IEEE Journal of Oceanic Engineering 49(3)Journal HALPDFDOI

20235 records

  1. Détecteur de Rao robuste pour un sonar à croix de Mills Olivier Lerda, Ammar Mian, Guillaume Ginolhac, Jean-Philippe Ovarlez GRETSIConference HALPDF
  2. New Robust Sparse Convolutional Coding Inversion Algorithm for Ground Penetrating Radar Images Matthieu Gallet, Ammar Mian, Guillaume Ginolhac, Esa Ollila, Nickolas Stelzenmuller IEEE Transactions on Geoscience and Remote SensingJournal HALPDFDOI
  3. Learning Graphical Factor Models with Riemannian Optimization Alexandre Hippert-Ferrer, Florent Bouchard, Ammar Mian, Titouan Vayer, Arnaud Breloy ECML PKDDConference HALPDFDOI
  4. Classification d'objets enfouis par un modèle du second d'ordre en utilisant des données GPR Douba Jafuno, Ammar Mian, Guillaume Ginolhac, Nickolas Stelzenmuller GRETSIConference HALPDF
  5. Optimisation Riemannienne pour l'apprentissage de graphes structurés Alexandre Hippert-Ferrer, Florent Bouchard, Ammar Mian, Titouan Vayer, Arnaud Breloy GRETSI, pp. 585–588Conference HALPDF

20224 records

  1. Nouvel algorithme d'inversion robuste pour le RADAR GPR Matthieu Gallet, Ammar Mian, Guillaume Ginolhac, Nickolas Stelzenmuller GRETSIConference HALPDF
  2. Riemannian classification of EEG signals with missing values Alexandre Hippert-Ferrer, Ammar Mian, Florent Bouchard, Frédéric Pascal EUSIPCOConference HALPDFDOI
  3. Classification of GPR Signals via Covariance Pooling on CNN Features within a Riemannian Framework Matthieu Gallet, Ammar Mian, Guillaume Ginolhac, Nickolas Stelzenmuller IGARSSConference HALPDFDOI
  4. Détection de changements basée sur les covariances de séries d’images SAR Ammar Mian, Arnaud Breloy, Frédéric Pascal Détection de changements et analyse des séries temporelles d’images 1: Méthodes non superviséesChapter HAL

20213 records

  1. An Overview of Covariance‐based Change Detection Methodologies in Multivariate SAR Image Time Series Ammar Mian, Guillaume Ginolhac, Jean-Philippe Ovarlez, Arnaud Breloy, Frédéric Pascal Change Detection and Image Time Series Analysis 1Chapter HALDOI
  2. A Comparative Study of Statistical-Based Change Detection Methods for Multidimensional and Multitemporal SAR Images Ammar Mian, Frédéric Pascal Advances in Condition Monitoring and Structural Health Monitoring, pp. 189–206Chapter HALDOI
  3. On-line Kronecker Product Structured Covariance Estimation with Riemannian geometry for t-distributed data Florent Bouchard, Arnaud Breloy, Ammar Mian, Guillaume Ginolhac EUSIPCO, pp. 856–859Conference HALPDFDOI

20202 records

  1. Robust Low-rank Change Detection for SAR Image Time Series Ammar Mian, Antoine Collas, Arnaud Breloy, Guillaume Ginolhac, Jean-Philippe Ovarlez IEEE JSTARS 13, pp. 3545–3556Journal HALPDFDOI
  2. Riemannian geometry for compound Gaussian distributions: Application to recursive change detection Florent Bouchard, Ammar Mian, Jialun Zhou, Salem Said, Guillaume Ginolhac, Yannick Berthoumieu Signal Processing 176, art. 107716Journal HALPDFDOI

20196 records

  1. Détection de Changement Robuste en Rang Faible pour les Séries Temporelles d'Images SAR Ammar Mian, Arnaud Breloy, Guillaume Ginolhac, Jean-Philippe Ovarlez GRETSIConference HALPDF
  2. New Robust Statistics for Change Detection in Time Series of Multivariate SAR Images Ammar Mian, Guillaume Ginolhac, Jean-Philippe Ovarlez, Abdourrahmane Atto IEEE Transactions on Signal Processing 67(2), pp. 520–534Journal HALPDFDOI
  3. Contributions to SAR Image Time Series Analysis Ammar Mian Thesis HALPDFDOI
  4. Design of New Wavelet Packets Adapted to High-Resolution SAR Images With an Application to Target Detection Ammar Mian, Jean-Philippe Ovarlez, Abdourrahmane Atto, Guillaume Ginolhac IEEE Transactions on Geoscience and Remote Sensing 57(6), pp. 3919–3932Journal HALPDFDOI
  5. Designing Sar Images Change-point Estimation Strategies Using an Mse Lower Bound Ammar Mian, Lucien Bacharach, Guillaume Ginolhac, Alexandre Renaux, Mohammed Nabil El Korso, Jean-Philippe Ovarlez ICASSP, pp. 5312–5316Conference HALPDFDOI
  6. Detection Methods Based on Structured Covariance Matrices for Multivariate SAR Images Processing Rayen Ben Abdallah, Ammar Mian, Arnaud Breloy, Abigaël Taylor, Mohammed Nabil El Korso, David Lautru IEEE Geoscience and Remote Sensing Letters 16(7), pp. 1160–1164Journal HALDOI

20182 records

  1. A Robust Change Detector for Highly Heterogeneous Multivariate Images Ammar Mian, Jean-Philippe Ovarlez, Guillaume Ginolhac, Abdourahmane M Atto ICASSPConference HALPDFDOI
  2. Robust Detection and Estimation of Change-Points in a Time Series of Multivariate Images Ammar Mian, Jean-Philippe Ovarlez, Guillaume Ginolhac, Abdourrahmane Atto EUSIPCOConference HALPDFDOI

20172 records

  1. Multivariate Change Detection on High Resolution Monovariate SAR Image Using Linear Time-Frequency Analysis Ammar Mian, Jean-Philippe Ovarlez, Guillaume Ginolhac, A Atto EUSIPCOConference HALPDFDOI
  2. Détection de changement sur images SAR monovariées par analyse temps-fréquence linéaire Ammar Mian, Jean-Philippe Ovarlez, Guillaume Ginolhac, Abdourrahmane Atto GRETSIConference HALPDF

3Students & tools

PhD — in progress

  • PhD · 2024—
    Matthieu VerlyndeApprentissage frugal pour la télédétectionCo-supervised with Y. Yan
  • PhD · 2025—
    Wenli ShiRéseau de diffusions sparses sur des données non textuellesCo-directors F. Pascal (L2S) & M. Jonckheere (LAAS)

PhD — defended

  • PhD · 2022—2025
    Douba JafunoClassification de signaux GPR — defended 4 Nov 2025Co-supervisors G. Ginolhac, N. Stelzenmuller
  • PhD · 2021—2024
    Olivier LerdaDétection robuste pour un sonar à croix de Mills — defended 17 Dec 2024Co-supervisors G. Ginolhac, J-P. Ovarlez

Master’s

  • M2 · 2025
    Thibault PautrelSPDnet federated learningCo-supervised with F. Bouchard and G. Ginolhac
  • M2 · 2025
    Thomas PetellatDétection de fissures volcaniques par apprentissage automatiqueCo-supervised with C. Lin-Kwong-Chon and Y. Yan
  • M2 · 2025
    Mamour NdiayePansharpening avec la variation totaleCo-supervised with Y. Mhiri
  • M2 · 2024
    Emma MolièrePansharpening and unrollingCo-supervised with A. Bhanot
  • M2 · 2024
    Matthieu VerlyndeEfficient algorithms for remote sensing
  • M2 · 2020
    Matthieu GalletInversion robuste de signaux GPRCo-supervised with G. Ginolhac

Software

  • Software
    QanatA command-line tool for managing and reproducing numerical experiments. Repository
  • Software
    Endless FieldAn endless procedural grassland rendered with WebGPU. You are the wind: carry a stream of petals, bloom glowing flowers, and watch the dry field turn green behind you. Live demoRepository

4Contact

Email
ammar.mian@univ-smb.fr
Office
LISTIC, Polytech Annecy-Chambéry, Université Savoie Mont-Blanc, Annecy
HAL
search results
GitHub
github.com/ammarmian

§On the building of this tower

Design
Ammar Mian
Drawn
Claude Opus 5in Claude Code
Stack
three.js 0.184 · TypeScript 5.6Vite 5.4 · anime.js 3.2
Type
Cormorant Garamond · Newsreader · JetBrains Mono · Pirata One
Source
HAL open archivethe shelves are fetched live, never a snapshot
Build
GitHub Actions → Pagesa static site; there is no server

Every room, every material and every tool here was chosen and specified by me. The code that builds them was written in conversation with Claude Opus 5.

There are no models and no textures anywhere in this site. Every shape in the tower is built from boxes and cylinders in code — the grain in the floorboards, the bark on the trees and the rust on the cars are all noise functions written into the shaders.