Working Student - Machine Learning

What you'll do

• Design and prototype ML models tailored for AR use cases under embedded constraints.
• Set up datasets and baselines relevant to AR tasks, defining evaluation metrics across accuracy, latency, memory usage, and energy.
• Implement and train deep learning models in PyTorch, including data pipelines and training loops that are easy to extend and reproduce.
• Explore efficiency techniques such as sparsity, pruning, quantization, or event-based representations to improve model performance.
• Profile models under embedded-like conditions using simulators, profiling tools, or edge accelerators to understand system-level behavior.

What we're looking for

• Currently enrolled in a Master’s program with relevant field experience.
• Strong background in linear algebra, probability, optimization, and deep learning fundamentals.
• Hands-on experience training deep learning models for computer vision using PyTorch or similar frameworks.
• Proficiency in Python and standard ML tooling including NumPy, Git, and experiment management.
• Interest in developing robust, reproducible codebases from research ideas.