Perception Engineer - Multimodal & Embodied AI (R6022)

What you'll do

• Evaluate models and methods against sensor data and autonomy needs, and build prototypes to assess performance and deployment potential.
• Combine learned and classical perception methods for detection, tracking, geometric vision, and sensor fusion.
• Optimize and integrate capabilities for edge deployment, contributing Python and C++ software and supporting adoption by other teams.

What they ask for

• PhD in computer vision, machine learning, robotics, or a related field, or an equivalent record of advanced research and hands-on engineering.
• Research depth in multimodal vision, visual foundation models, video understanding, open-vocabulary perception, embodied learning, or a related area.
• Strong Python skills and experience with PyTorch or a comparable ML framework.
• Experience building substantial research software; interest in production-quality C++ development.

Nice to have

• Experience moving a research prototype into sustained use, or working with robotics, real sensors, temporal data, or multi-sensor perception.
• Familiarity with vision-language models, vision-language-action models, embodied AI, C++, edge inference, or model optimization tools such as ONNX, TensorRT, or CUDA.

Tools

Python, PyTorch, C++, ONNX, TensorRT, CUDA

AI focus

Transfers multimodal computer vision, visual foundation models, and embodied AI research into perception capabilities for autonomous aircraft and vehicles, including edge deployment.