Computer Vision Intern

At Harmattan AI, we are a next-generation defense prime building autonomous and scalable defense systems. Driven by rigorous engineering developments of new defense products based on recent robotics and AI developments, we are on a steep growth trajectory. If you are interested in a career in a highly technical environment, thrive on pushing boundaries, and want to achieve ambitious goals, we would love to hear from you.

Harmattan AI is operating a drone-integrated Synthetic Aperture RADAR (SAR) system designed for Intelligence, Surveillance, and Reconnaissance (ISR) applications. This system generates high-resolution RADAR imagery over extended geographical areas. Currently, the primary challenge is the human-intensive nature of image interpretation. While the system provides raw and derived image products, the identification of key features, targets, or changes of interest is left to the user.

The goal of this internship is to bridge the gap between raw data generation and actionable intelligence by creating an intelligent, automated analysis layer.

You will be responsible for developing a software overlay and processing pipeline designed to accelerate and facilitate SAR image interpretation. This involves researching, implementing, and validating computer vision and image processing techniques tailored for SAR data characteristics.

The core mission is to rapidly and reliably identify, characterize, and highlight points of interest (POI) across large SAR images, providing the user with a focused and effective visualization interface.

• Literature review: Conduct a comprehensive review of modern target detection techniques applicable to SAR imagery

• Algorithm Implementation: Develop and implement POI detection and characterization algorithms. The features to be developed include:

• Change Detection POI identification: Implement algorithms to automatically detect and delineate moved objects or significant variations visualized in Amplitude Change Detection (ACD) products.

• Infrastructure Detection: Develop and train models to identify man-made infrastructure (e.g., roads, buildings, power lines) obscured or blending into complex natural environments (vegetation, forest canopy) using Polarimetric or Multi-Temporal Coherence (MTC) data
• Vehicle Detection and measurement: Implement a pipeline to detect stationary vehicles on the SAR image and develop geometric estimation techniques to determine their dimensions (length, width) in ground units
• Documentation: Document all work performed, including methodologies, results, performances and conclusions.
• Flight Tests Participation: Actively participate in flight test sessions to gather real-world data and validate the effectiveness of the developed algorithms in operational conditions.

• Educational Background: Graduate student in an Engineering School, Master's program or equivalent, seeking a Final Year Internship or a Gap Year Internship. Specialization in Computer Science, Data Science, Signal and Image Processing or Remote Sensing
• Technical skills: Strong proficiency in Python (including numerical libraries, e.g. NumPy, Pandas, OpenCV,...) or C++. Familiarity with Image processing concepts and Computer Vision. Knowledge in Deep Learning is a plus. Knowledge of RADAR principles would be greatly appreciated but are not mandatory.
• Mindset: You are curious to learn, autonomous and able to take initiative.
• Communication: Excellent written and verbal communication. You are able to clearly articulate and present your work.

We look forward to hearing how you can help shape the future of autonomous defense systems at Harmattan AI.