I am interested in computer vision with a focus on unsupervised image collection analysis (ERC DISCOVER), and applications to historical data (also ANR EnHerit and ANR VHS and ANR EIDA) and Earth imagery analysis. Contact me if you are interested in historical data or earth imagery for M2/PhD ( 2023 internship offers ), research engineer and postoc positions.
Mathieu Aubry
senior researcher
Contact
Imagine team | LIGM lab | ENPC, École des Ponts ParisTechIMAGINE/LIGM
École des Ponts ParisTech
6-8, Av Blaise Pascal - Cité Descartes
Champs-sur-Marne
77455 Marne-la-Vallée cedex 2
email: mathieu [dot] aubry [at] imagine [dot] enpc [dot] fr
phone: +33 (0)1 64 15 21 74
Research interests
Students
Current PhD students:- Tom Monnier
- Ioannis Siglidis
- Elliot Vincent
- Sonat Baltaci
- Malamatenia Vlachou Efstathiou; (in History with Dominique Stuzmann at IRHT)
- Yann Labbé (at INRIA/ENS with Justin Caprentier and Josef Sivic)
- Robin Champenois (at ENS in Art with Béatrice Joyeux-Prunel)
- Romain Loiseau (with Loic Landrieu)
- Francisco Massa (with Renaud Marlet, now at Meta)
- Vianney Loing (in Mechnaical Engineering, with Jean-François Caron, Olivier Baverel and Renaud Marlet, now athttps://arcure.net/" target="_blank">Arcure)
- Théophile Dalens (at INRIA with Josef Sivic, now at Grabango)
- Thibault Groueix (now at Adobe research)
- Othman Sbaï (in CIFRE at FAIR with Camille Couprie, now at Meta)
- Xi Shen (now at Tecent research)
- Francois Darmon (with Pascal Monasse, transitioning to Meta)
- Théo Deprelle (now at Veesual)
- Simon Roburin (in CIFRE at Valeo AI with Patrick Perez and Renaud Marlet)
Past Post-doc: Nicolas Gonthier (now IPEF working at IGN)
Publications
Also see Google Scholar
Learnable Earth Parser: Discovering 3D Prototypes in Aerial Scans
ArXiv 2023
The Learnable Typewriter A Generative Approach to Text Line Analysis
ArXiv 2023
MegaPose: 6D Pose Estimation of Novel Objects via Render & Compare
CoRL 2022
A Model You Can Hear: Audio Classification with Playable Prototypes
ISMIR 2022, CVPRw 2022
Share With Thy Neighbors: Single-View Reconstruction by Cross-Instance Consistency
ECCV 2022
Online Segmentation of LiDAR Sequences: Dataset and Algorithm
ECCV 2022
NeuralWarp: Improving neural implicit surfaces geometry with patch warping
CVPR 2022
Focal Length and Object Pose Estimation via Render and Compare
CVPR 2022
Spatially-consistent Feature Matching and Learning for Heritage Image Analysis
IJCV 2022
Re-ranking for image retrieval and transductive few-shot classification
NeurIPS 2021
Learning Co-segmentation by Segment Swapping for Retrieval and Discovery
ArXiv 2021, CVPR 2022 workshops
Representing Shape Collections with Alignment-Aware Linear Models
3DV 2021
Unsupervised Layered Image Decomposition into Object Prototypes
ICCV 2021
Image Collation: Matching illustrations in manuscripts
ICDAR 2021
Single-view robot pose and joint angle estimation via render & compare
CVPR 2021
Spherical Perspective on Learning with Batch Norm
ArXiv 2020, Neurocomputing 2022
docExtractor: An off-the-shelf historical document element extraction
ICFHR 2020
CosyPose: Consistent multi-view multi-object 6D pose estimation
ECCV 2020
Impact of base dataset design on few-shot image classification
ECCV 2020
Monte-Carlo Tree Search for Efficient Visually Guided Rearrangement Planning
RAL 2020
Large-Scale Historical Watermark Recognition:dataset and a new consistency-based approach
ICPR 2020
Download pdf | View project web page | Web App | paper on Web App
Learning elementary structures for 3D shape generation and matching
NeurIPS 2019
Unsupervised Image Decomposition in Vector Layers
ArXiv 2019 / ICIP 2020
Unsupervised cycle-consistent deformation for shape matching
SGP 2019
Discovering Visual Patterns in Art Collections with Spatially-consistent Feature Learning
CVPR 2019
Virtual Training for a Real Application: Accurate Object-Robot Relative Localization without Calibration
IJCV 2018
AtlasNet: A Papier-Mâché Approach to Learning Surface Generation
CVPR 2018
3D Sketching using Multi-View Deep Volumetric Prediction
I3D 2018
Crafting a multi-task CNN for viewpoint estimation
BMVC 2016
Automatic 3D Car Model Alignment for Mixed Image-Based Rendering
3DV 2016
Deep Exemplar 2D-3D Detection by Adapting from Real to Rendered Views
CVPR 2016
Learning Dense Correspondence via 3D-guided Cycle Consistency
CVPR, 2016
Dex-Net 1.0: A Cloud-Based Network of 3D Objects for Robust Grasp Planning Using a Multi-Armed Bandit Model with Correlated Rewards
ICRA, 2016
Understanding deep features with computer-generated imagery
ICCV, 2015
Download pdf (ArXiv preprint) | bibtex | View project web page
Visual geo-localization of non-photographic depictions via 2D-3D alignment
Springer book chapter on Visual Analysis and Geo-Localization of Large Scale Imagery, 2015
Anisotropic Laplace-Beltrami Operators for Shape Analysis
In Sixth Workshop on Non-Rigid Shape Analysis and Deformable Image Alignment (NORDIA), 2014
Seeing 3D chairs: exemplar part-based 2D-3D alignment using a large dataset of CAD models
CVPR, 2014
Painting-to-3D Model Alignment Via Discriminative Visual Elements
ACM Transactions on Graphics, 2014 (will be presented at SIGGRAPH 2014)
Fast Local Laplacian Filters: Theory and Applications
ACM Transactions on Graphics, 2014 (will be presented at SIGGRAPH 2014)
A Super-resolution Framework for High-Accuracy Multiview Reconstruction
International Journal of Computer Vision, volume 106, 2014
Decoupling Photometry and Geometry in Dense Variational Camera Calibration
IEEE International Conference on Computer Vision (ICCV), 2011
Pose-Consistent 3D Shape Segmentation Based on a Quantum Mechanical Feature Descriptor
Pattern Recognition (Proc. DAGM), Springer, 2011
The Wave Kernel Signature: A Quantum Mechanical Approach To Shape Analysis
IEEE International Conference on Computer Vision (ICCV) - Workshop on Dynamic Shape Capture and Analysis (4DMOD), 2011
Teaching
- ENPC Deep Learning (M1/M2, since 2021)
- MVA 3D reconstruction, Object Recognition and Deep Learning (M2, one lecture in each)
- ENS Introduction to Computer Vision (M1/M2, 2017-2022)
- ENPC Computer Vision for Mechanics of materials (L3, 2015-2021)
Short CV
-
Aug 2019:Habilitation à diriger des recherches (HDR, kind of second PhD in France) on "Neural Networks for Cross-Modal Recognition and Vector Object Generation".
-
Sept 2014-now:Permanent researcher in the Imagine team in the École des Ponts ParisTech (administratively IPEF since 2011 and ICPEF since 2021)
-
Jan 2015-Dec 2015:Visiting Alexei Efros at UC Berkeley (a.k.a. postdoc).
-
May 2015 :PhD on "Representing 3D models for alignment and recognition".
-
Sept 2011-Sept 2014:PhD student, co-advised by Josef Sivic and Daniel Cremers.
-
Feb 2011-Sept 2011:Intern at Adobe, co-advised by Sylvain Paris and Frédo Durand.
-
Winter term 2010:Master thesis at the Technical University of Munich on camera calibration and 3D recognition. Work done with Daniel Cremers, Bastian Goldluecke and Ulrich Schlickewei.
-
2009-2010:I followed a general program in the ENPC and graduated in the Mathematics, Vision and Learning (MVA) master of the ENS-Cachan.
-
Summer term 2009:Research project on the foundations and the understanding of mathematics with G. Lakoff in UC Berkeley (research internship prize of the École Polytéchnique).
-
2006-2009:generic Master in Mathematics, Physics and Cognitive Sciences at the École Polytéchnique.
I was/am area chair (AC) for CVPR18 and CVPR20, MVA 2021, BMVC 2021, ACCV2021, ACCV2022, CVPR 2023, ICCV 2023, NeurIPS 2023. I was in the award comitee for CVPR2022. I was amongst the top rated reviewers for CVPR19, NeurIPS19 and ECCV 2022. I was associated edidor for CVIU from 2021 to 2023.
Interns and Master students: Karan Dwivedi, Pierre-Alain Langlois, Thomas Belos, David Tidmarsh, Théo Deprelle, Sophie Bodenes, Oumayma Bounou, Tom Monnier, Elliot Vincent, Ryad Kaoua, Van Nguyen Nguyen, Virginie Loison, Xiwei Hu, Ioannis Siglidis, Julien Gaubil.
For more detailed informations, you can download a full version of my CV (might not be up to date) .