Day-to-Day Responsibilities
Collect, clean and label large sets of images or video frames, and check that the data is accurate and varied.
Design and train deep learning models, such as convolutional neural networks, for tasks like object detection, segmentation and tracking.
Apply image processing methods such as filtering, colour correction and camera calibration to prepare inputs.
Test models in real-world conditions such as low light, motion blur and partial occlusion, and measure accuracy and speed.
Optimise models to run on limited hardware such as cameras, phones or vehicle computers.
Deploy models into products through APIs or edge systems, then monitor their performance after launch.
Work with product managers, hardware teams and domain experts such as radiologists or road-safety engineers.


