Fatemeh N. Nokabadi

I am a Machine Learning Scientist specializing in computer vision and modern AI systems. My work focuses on building reliable, high-impact models that perform in real-world conditions, not just benchmarks.

I hold a Ph.D. in Electrical Engineering from Université Laval and Mila, supervised by Prof. Christian Gagné and Prof. Jean-François Lalonde. My doctoral research focused on adversarial robustness of learning-based object trackers; designing attack frameworks that expose critical vulnerabilities in state-of-the-art transformer trackers.

Prior to my PhD, I conducted research on single-object tracking via deep reinforcement learning at LVSN, Université Laval, and built computer vision systems for real-world industry applications.

I am particularly interested in vision-language models, structured understanding of visual data, and robustness of learning systems under challenging environments.

• Built an end-to-end system for extracting structured information from complex documents using VLMs.
• Handled diverse data types and real-world variability at scale.
• Delivered reliable performance in production workflows.

• Developed frameworks to evaluate failure modes of modern vision models.
• Exposed critical weaknesses under adversarial conditions.
• Improved understanding of model reliability in safety-critical scenarios.

Transactions on Machine Learning Research 2024 · Proceedings of the 22nd Conference on Robots and Vision 2025 · Journal-to-conference track & AdvML Frontiers at NeurIPS2024

• Developed models for disentangled representation learning in video data.
• Separated motion and appearance factors in complex scenes.
• Improved downstream tasks like segmentation and temporal reasoning.

WiML at ICML 2024

• Built models for fine-grained visual verification tasks.
• Leveraged subtle visual cues for authenticity detection.
• Achieved high accuracy in real-world scenarios.

• Developed graph-based saliency detection algorithms using local, global, and mixed feature representations.
• Proposed multiple methods including random graph and global contrast graph approaches for salient object detection in images.

Multimedia Tools and Applications 2018 · Signal, Image and Video Processing 2018 · SPIS 2015

• Researched and implemented real-time multi-object tracking algorithms for cluttered video frames, incorporating salient object detection to prioritize targets.
• Designed and deployed a MATLAB and C++ tracking system using OpenCV for moving-camera video streams, improving tracking speed by 30%.