AI & Computer Vision Intern - Data augmentation
About Us
Harmattan AI is a next-generation defense prime building autonomous and scalable defense systems. Following the close of a $200M Series B, valuing the company at $1.4 billion, we are expanding our teams and capabilities to deliver mission-critical systems to allied forces.
Our work is guided by clear values: building technologies with real-world impact, pursuing excellence in everything we do, setting ambitious goals, and taking on the hardest technical challenges. We operate in a demanding environment where rigor, ownership, and execution are expected.
About the Role
To train highly robust perception systems, we need aerial imagery spanning a massive variety of conditions. You will engineer an advanced Generative AI pipeline capable of completely transforming the context of existing datasets; shifting time-of-day (day to night), changing seasons (summer to winter), or altering entire biomes and weather systems; while perfectly preserving small, critical target objects like drones.
Refine the Generative Architecture: Take ownership of a sophisticated multi-pass diffusion pipeline (structural background replacement + lighting/atmospheric glazing) to seamlessly adapt scene contexts while maximizing physical realism.
Solve Edge Cases: Improve custom masking and high-res depth-patching algorithms (e.g., histogram matching, seamless blending) to anchor small objects in 3D space, eliminating generative artifacts and "sticker" effects.
Scale & Validate: Generate large-scale augmented datasets and rigorously quantify their impact on downstream model performance. Design experiments to measure how the inclusion and varying ratios of this synthetic data directly improve the accuracy, recall, and robustness of object detectors (e.g., YOLO) when tested against real-world edge cases.
Requirements
Education: Currently pursuing or recently completed a Master’s degree in Computer Science, Robotics, Electrical Engineering, or a related field with a focus on Machine Learning.
Deep Learning: Strong understanding of CNN architectures, object detection frameworks, and modern loss functions, as well as the tracking world and its problematics.
Software Engineering: Proficiency in Python (PyTorch/TensorFlow) and comfortable working in C++.
Linux/Embedded: Experience working in a Linux environment; familiarity with Git is a plus.
Problem Solving: A rigorous approach to debugging and an "engineering first" mindset, valuing performance over theoretical complexity.
Language: Fluency in English; French is a plus.
Bonus
Experience with NVIDIA Jetson platforms and hardware-accelerated inference.
FPV pilot experience or hobbyist interest in UAVs.
Previous experience with synthetic data generation (e.g., NVIDIA Isaac Sim, Gazebo).
We look forward to hearing how you can help shape the future of autonomous defense systems at Harmattan AI.