Hughes Perreault
AI Researcher · Institut de recherche d'Hydro-Québec (IREQ)
About
I am an AI researcher at the Institut de recherche d'Hydro-Québec (IREQ), where I work on
machine learning for the energy sector. My research centers on computer vision and efficient
deep learning: detecting, segmenting and tracking objects in images and videos, and fusing
information from multiple sensors. Before joining IREQ, I was Principal AI Developer at Haylem
and an AI Scientist at Thales, where I developed multimodal image fusion, thermal and visible
image registration, and lightweight target detection for edge devices. I received my Ph.D. in
computer engineering from Polytechnique Montréal in 2021, with a thesis on the detection and
segmentation of road users in images and videos.
Fun fact, my Dijkstra number is five (click here) and my Erdos number is also five (click
here).
Research Interests
- Computer vision: object detection, instance segmentation and tracking
- Multimodal image fusion and registration (thermal and visible)
- Efficient deep learning for edge devices
- Machine learning for the energy sector
News
- Jan 2026 Joined the Institut de recherche d'Hydro-Québec (IREQ) as an AI Researcher.
- Mar 2025 Joined Haylem as Principal AI Developer.
- Aug 2023 Our paper on lightweight target detection was presented at GVSETS 2023 (arXiv).
- Jun 2023 MobileFuse published at FUSION 2023.
- Oct 2021 Joined Thales as an AI Scientist.
Positions
Selected Publications
All Publications
2023
- A Lightweight Convolutional Neural Network for Target Detection in an Edge Environment
2023 NDIA Michigan Chapter Ground Vehicle Systems Engineering and Technology Symposium (GVSETS)
@inproceedings{perreault2023lightweight, title = {A Lightweight Convolutional Neural Network for Target Detection in an Edge Environment}, author = {Perreault, Hughes and Ahmad, Ola and Czuboka, Daniel}, booktitle = {2023 NDIA Michigan Chapter Ground Vehicle Systems Engineering and Technology Symposium (GVSETS)}, year = {2023}, eprint = {2611.09756}, archivePrefix = {arXiv} } - MobileFuse: Multimodal Image Fusion at the Edge
2023 26th International Conference on Information Fusion (FUSION)
@inproceedings{perreault2023mobilefuse, title = {{MobileFuse}: Multimodal Image Fusion at the Edge}, author = {Perreault, Hughes and Debaque, Benoit and David, Rares and Drouin, Marc-Antoine and Duclos-Hindi{\'e}, Nicolas and Roy, Simon}, booktitle = {2023 26th International Conference on Information Fusion (FUSION)}, pages = {1--8}, year = {2023}, doi = {10.23919/FUSION52260.2023.10224104} }
2022
- Multimodal Deep Homography Estimation Using a Domain Adaptation Generative Adversarial Network
2nd Workshop on Multimodal Big Data (MMBD 2022), 2022 IEEE International Conference on Big Data (Big Data)
@inproceedings{pouplin2022multimodal, title = {Multimodal Deep Homography Estimation Using a Domain Adaptation Generative Adversarial Network}, author = {Pouplin, Thomas and Perreault, Hughes and Debaque, Benoit and Drouin, Marc-Antoine and Duclos-Hindie, Nicolas and Roy, Simon}, booktitle = {2022 IEEE International Conference on Big Data (Big Data)}, pages = {3635--3641}, year = {2022}, doi = {10.1109/BigData55660.2022.10020557} } - Thermal and Visible Image Registration Using Deep Homography
2022 25th International Conference on Information Fusion (FUSION)
@inproceedings{debaque2022thermal, title = {Thermal and Visible Image Registration Using Deep Homography}, author = {Debaque, Benoit and Perreault, Hughes and Mercier, Jean-Philippe and Drouin, Marc-Antoine and David, Rares and Chatelais, Benedicte and Duclos-Hindie, Nicolas and Roy, Simon}, booktitle = {2022 25th International Conference on Information Fusion (FUSION)}, pages = {1--8}, year = {2022}, doi = {10.23919/FUSION49751.2022.9841256} }
2021
- Détection et segmentation des usagers de la route dans des images et vidéos
Ph.D. thesis, Polytechnique Montréal
@phdthesis{perreault2021these, title = {D{\'e}tection et segmentation des usagers de la route dans des images et vid{\'e}os}, author = {Perreault, Hughes}, school = {Polytechnique Montr{\'e}al}, year = {2021}, url = {https://publications.polymtl.ca/9471/} } - PolyTrack: Tracking with Bounding Polygons
Machine Learning for Autonomous Driving Workshop (ML4AD), NeurIPS 2021
@inproceedings{faure2021polytrack, title = {{PolyTrack}: Tracking with Bounding Polygons}, author = {Faure, Gaspar and Perreault, Hughes and Bilodeau, Guillaume-Alexandre and Saunier, Nicolas}, booktitle = {Machine Learning for Autonomous Driving Workshop (ML4AD), NeurIPS}, year = {2021}, eprint = {2111.01606}, archivePrefix = {arXiv} } - CenterPoly: Real-time Instance Segmentation Using Bounding Polygons
Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops (ICCVW)
@inproceedings{perreault2021centerpoly, title = {{CenterPoly}: Real-time Instance Segmentation Using Bounding Polygons}, author = {Perreault, Hughes and Bilodeau, Guillaume-Alexandre and Saunier, Nicolas and H{\'e}ritier, Maguelonne}, booktitle = {2021 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW)}, pages = {2982--2991}, year = {2021}, doi = {10.1109/ICCVW54120.2021.00333} } - FFAVOD: Feature Fusion Architecture for Video Object Detection
Pattern Recognition Letters, vol. 151
@article{perreault2021ffavod, title = {{FFAVOD}: Feature Fusion Architecture for Video Object Detection}, author = {Perreault, Hughes and Bilodeau, Guillaume-Alexandre and Saunier, Nicolas and H{\'e}ritier, Maguelonne}, journal = {Pattern Recognition Letters}, volume = {151}, pages = {294--301}, year = {2021}, doi = {10.1016/j.patrec.2021.09.002} }
2020
- SpotNet: Self-Attention Multi-Task Network for Object Detection
17th Conference on Computer and Robot Vision (CRV)
@inproceedings{perreault2020spotnet, title = {{SpotNet}: Self-Attention Multi-Task Network for Object Detection}, author = {Perreault, Hughes and Bilodeau, Guillaume-Alexandre and Saunier, Nicolas and H{\'e}ritier, Maguelonne}, booktitle = {2020 17th Conference on Computer and Robot Vision (CRV)}, pages = {230--237}, year = {2020}, doi = {10.1109/CRV50864.2020.00038} } - RN-VID: A Feature Fusion Architecture for Video Object Detection
International Conference on Image Analysis and Recognition (ICIAR 2020), LNCS
@inproceedings{perreault2020rnvid, title = {{RN-VID}: A Feature Fusion Architecture for Video Object Detection}, author = {Perreault, Hughes and H{\'e}ritier, Maguelonne and Gravel, Pierre and Bilodeau, Guillaume-Alexandre and Saunier, Nicolas}, booktitle = {Image Analysis and Recognition (ICIAR 2020)}, series = {Lecture Notes in Computer Science}, pages = {125--138}, publisher = {Springer}, year = {2020}, doi = {10.1007/978-3-030-50347-5_12} }
2019
- Road User Detection in Videos
arXiv preprint
@misc{perreault2019roaduser, title = {Road User Detection in Videos}, author = {Perreault, Hughes and Bilodeau, Guillaume-Alexandre and Saunier, Nicolas and Gravel, Pierre}, year = {2019}, eprint = {1903.12049}, archivePrefix = {arXiv} }