Deep learning · 3D perception · robotics
William Guimont-Martin
Québec, Canada
I am a doctoral researcher in computer science at Université Laval. My research focuses on deep learning and 3D sensors for autonomous vehicles, with broader interests in point clouds, LiDAR, computer vision, and mobile robotics.
Education
Degrees
Ph.D. in Computer Science
Université Laval, Québec · 2023–present
Research in deep learning and 3D sensors for autonomous vehicles.
M.Sc. in Computer Science, thesis option
Université Laval, Québec · 2021–2023
Research in deep learning and 3D sensors for autonomous vehicles.
B.Eng. in Software Engineering, Distinction Profile
Université Laval, Québec · 2017–2021
Scholarships and honours
Graduate and doctoral awards
EGGENIUS Entrepreneurship Scholarship — Hope
Claude Dussault · 2024–2025 · $1,000
Awarded to encourage students in the Faculty of Science and Engineering to consider entrepreneurship as a professional path and to recognize strong entrepreneurial potential.
Canada Graduate Scholarship — Doctoral (CGS D)
Natural Sciences and Engineering Research Council of Canada (NSERC) · 2023 · $35,000 per year for three years
Awarded to high-calibre doctoral students in recognition of academic excellence, research potential, and interpersonal skills.
Postgraduate Scholarship — Doctoral (PGS D)
Natural Sciences and Engineering Research Council of Canada (NSERC) · 2023 · $21,000 per year for three years
This scholarship was converted by NSERC into the CGS D award. It recognizes academic excellence, research potential, and interpersonal skills.
Doctoral Research Scholarship
Fonds de recherche du Québec — Nature et technologies (FRQNT) · 2023 · $25,000 per year for four years
Awarded in recognition of excellence in academic research. Award record.
Alexander Graham Bell Canada Graduate Scholarship
Natural Sciences and Engineering Research Council of Canada (NSERC) · 2021 · $17,500
Awarded to high-calibre master’s students in recognition of academic excellence, research potential, and interpersonal skills.
Master’s Research Scholarship
Fonds de recherche du Québec — Nature et technologies (FRQNT) · 2021 · $17,500 per year for two years
Awarded in recognition of excellence in academic research. Award record.
Pierre-Marchand Excellence Scholarship — Academic Results, Graduate Level
Pierre Marchand · 2021 · $2,000
Recognized academic excellence among students enrolled in the master’s program in computer science at Université Laval.
Excellence Scholarship in Computing and Electrical Engineering Fields
Government of Québec, Ministère de l’Enseignement supérieur · 2021 · $1,000
Awarded for academic excellence following graduation from Université Laval’s software engineering program.
Undergraduate awards
Undergraduate Student Research Award
Natural Sciences and Engineering Research Council of Canada (NSERC) · 2020 · $4,500
Supported a research internship at Université Laval’s Northern Robotics Laboratory.
FRQNT Supplement to the NSERC Undergraduate Student Research Award
Fonds de recherche du Québec — Nature et technologies · 2020 · $1,500
Supplemented the NSERC award for a research internship at Norlab. Award record.
Industrial Undergraduate Student Research Award
Natural Sciences and Engineering Research Council of Canada (NSERC) · 2018 · $4,500
Supported an industrial research internship at Olympus NDT Canada.
FRQNT Industrial Experience Supplement
Fonds de recherche du Québec — Nature et technologies · 2018 · $2,000
Supplemented the NSERC industrial research award for work at Olympus NDT Canada. Award record.
Leadership and Sustainable Development Scholarship — Scientific Leadership
Fondation Famille Choquette · 2017–2021 · $8,000 over four years
Recognized exceptional scientific leadership, substantial community involvement, and academic excellence. The award highlighted my volunteer work as a mentor for a FIRST Robotics Competition team.
Admission Excellence Scholarship
Hydro-Québec · 2017 · $3,000
Awarded for academic excellence upon admission to the software engineering bachelor’s program.
Software Engineering Admission Scholarship
Faculty of Science and Engineering, Université Laval · 2017 · $1,500
One of four admission scholarships awarded to incoming software engineering students.
Experience
Teaching
Teaching Assistant — Deep Learning
Université Laval, Québec · Fall 2021; Winter 2023, 2024, and 2025
GLO-4030 and GLO-7030, Deep Learning. Responsible for laboratories, grading, and the design of teaching material. In Winter 2025, I took responsibility for the entire course apart from lectures.
Teaching Assistant — Mobile Robotics
Université Laval, Québec · Winter 2022; Fall 2022, 2023, and 2024
GLO-4001 and GLO-7021, Introduction to Mobile Robotics. Responsible for laboratories, grading, and the design of teaching material.
Teaching Assistant — Multidisciplinary Design Project
Université Laval, Québec · Fall 2023
GLO-3013, Multidisciplinary Design Project. Helped redesign the course and developed ROS 2-based autonomous systems using LiDAR sensors.
Teaching Assistant — Operating Systems
Université Laval, Québec · Summer 2023
GLO-2001 and IFT-2001, Operating Systems for Engineering and Operating Systems. Created new teaching material on Bash and Docker.
Teaching Assistant — Introduction to Software Engineering Processes
Université Laval, Québec · Winter 2021
GLO-2003, Introduction to Software Engineering Processes. Graded practical assignments, mentored students, and designed teaching material.
Research and industry
Deep Learning Research Intern
Norlab, Université Laval, Québec · Summer and Fall 2020
Researched the application of deep neural networks to LiDAR data and point clouds for object detection, including solid-state and full-waveform LiDAR.
Virtual and Augmented Reality Research Intern
Bentley Systems, Québec · Summer 2019
Researched virtual and augmented reality technologies for infrastructure management. Developed computer vision and photogrammetry algorithms that contributed to a U.S. patent.
Industrial Research Intern
Olympus NDT Canada, Québec · Summer 2018
Researched automated ultrasonic-data analysis and visualization. Developed computer vision algorithms for new products.
Research and Development Developer
Devoray, Québec · Summer 2015; Winter 2016 and 2017
Developed computer vision algorithms and automated industrial production lines.
Publications and research contributions
See the dedicated Papers page for publication and source links.
Research publications
MaskBEV: Joint Object Detection and Footprint Completion for Bird’s-Eye View 3D Point Clouds
Guimont-Martin, W., Fortin, J.-M., Pomerleau, F., & Giguère, P. (2023). IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).
I introduced MaskBEV, a neural architecture for object detection in bird’s-eye-view 3D point clouds. It predicts object-footprint masks rather than bounding boxes, enabling completion under significant occlusion and reformulating detection as classification rather than regression. The approach was evaluated on SemanticKITTI and KITTI.
SilvaScenes: Tree Segmentation and Species Classification from Under-Canopy Images in Natural Forests
Duclos, D. A., Guimont-Martin, W., Jeanson, G., Larochelle-Tremblay, A., Defosse, T., Moore, F., et al. (2025). arXiv:2510.09458.
SilvaScenes introduces an instance-segmentation and tree-species classification dataset collected under forest canopies across five Québec bioclimatic domains. It contains 1,476 expert-annotated trees from 24 species and captures the occlusion and lighting variability of natural forests.
I implemented the segmentation models and contributed to the paper, experiments, result analysis, and figure design.
Using Citizen Science Data as Pre-Training for Semantic Segmentation of High-Resolution UAV Images for Natural Forests Post-Disturbance Assessment
Nasiri, K., Guimont-Martin, W., LaRocque, D., Jeanson, G., Bellemare-Vallières, H., Grondin, V., et al. (2025). Forests, 16(4), 616.
This work uses citizen-science data from iNaturalist to pre-train semantic segmentation models for UAV forest imagery. A sliding-window classifier generates pseudo-labels from unannotated drone images, reducing dependence on expensive manual annotation.
I played a key role in project management and coordination, supervised the master’s student during the principal investigator’s absence, designed and ran experiments, analyzed results, prepared figures, and contributed substantially to the paper.
Proprioception Is All You Need: Terrain Classification for Boreal Forests
LaRocque, D., Guimont-Martin, W., Duclos, D. A., Giguère, P., & Pomerleau, F. (2024). IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).
The work introduces BorealTC, a proprioceptive terrain-classification dataset combining IMU, odometry, and motor-current measurements for surfaces including snow, ice, and silty loam.
I contributed substantially to the project’s conception and management, led the implementation of CNN-, RNN-, and Transformer-based terrain classifiers, helped design and run experiments, analyzed results, and contributed significantly to the writing.
Replication Study and Benchmarking of Real-Time Object Detection Models
Asselin, P. L., Coulombe, V., Guimont-Martin, W., & Larrivée-Hardy, W. (2024). arXiv:2405.06911.
This study evaluates the reproducibility and accuracy-speed trade-off of YOLOv7, RTMDet, ViTDet, and DETR on MS COCO 2017. We built a unified training and evaluation pipeline for fair comparisons across several hardware architectures and found that anchor-free models offered the strongest accuracy-speed trade-off.
I contributed equally to the model implementations and evaluation pipeline and served as the paper’s lead writer.
Educational contribution
Teaching Software Development Skills: A Workshop-Based Approach for an Undergraduate Program
Hardy, S. V. (2025). Proceedings of the Canadian Engineering Education Association (CEEA).
I developed modules on Bash scripting and task automation, Docker containerization, code refactoring, Clean Code principles, and test-driven development through programming katas. I had substantial freedom in the pedagogical design, emphasizing durable principles over individual tools.
Patent
Aerial Cable Detection and 3D Modeling from Images
Côté, Stéphane, and William Guimont-Martin. U.S. Patent No. 11,521,357. Issued December 6, 2022.
Developed during my 2019 internship at Bentley Systems, the patent covers the detection and 3D modelling of aerial cables from images. I designed and evaluated the computer vision and photogrammetry algorithm used to identify power, telephone, and suspension lines and generate their 3D digital-twin representation. I also contributed to dataset collection, performance analysis, and patent drafting with Bentley Systems’ ARLab director.
Conference
- IEEE/RSJ International Conference on Intelligent Robots and Systems — IROS 2023
Community involvement
Robotics Mentor
Team 5440 Les Chevaliers, École secondaire de La Seigneurie · 2016–present
I mentor secondary-school students in robotics and programming, supporting the team from its creation through its growth. The role has developed my leadership, multidisciplinary project-management skills, and ability to coordinate a group toward a common objective. The team participated in FIRST Robotics from 2016 through 2025 and moved to the CRC competition for the 2026 season.
Additional training
Courses and certifications
Leica BLK2GO and Cyclone Register 360
Summer 2024 · 16 hours
Training in BLK2GO scanning and mapping with Cyclone Register 360.
Planning a Coherent and Engaging Course: University Pedagogy
Winter 2024 · DVP-8203, Faculty of Graduate and Postdoctoral Studies
Completed an applied course-design project for GLO-4030 and GLO-7030, Deep Learning. The proposed improvements were implemented in Winter 2025, resulting in stronger evaluations and better pedagogical alignment.
Safe Snowmobile Operation
Winter 2024 · SST-35
Training in snowmobile trip planning, basic mechanical repairs, and safe operation.