Computer Vision & ROS2 Engineer for Robotics Project

Looking for an experienced computer vision engineer to join an ongoing robotics project. You'll be working on a real system, not a demo, so reliability and clean code matter more than fancy models. What you'll be working on - Reading and processing video frames from MCAP files using the mcap Python library - Decoding H.264 video frames to numpy arrays for AI inference - Building and improving object detection and damage detection pipelines - OCR pipeline with a two-stage approach (fast frame selection + heavy model inference) - Integrating vision pipeline with ROS2 topics - Working with Jetson edge hardware eventually Skills required - Strong Python - Computer vision with OpenCV and numpy - Object detection models (YOLO, RF-DETR or similar) - OCR pipelines - Basic ROS2 knowledge or willingness to learn quickly - Experience with video processing and frame extraction - FastAPI for backend endpoints Nice to have - MCAP format familiarity - Jetson or edge device experience - GStreamer or WebRTC knowledge What to expect - Active project with a real hardware team - Clear technical direction - Ongoing work if things go well