AI engineer in computer vision and deep learning. I cover the full lifecycle, from data preparation and annotation to training, optimization and deployment in real industrial settings.
Step through the stages. The plate changes as it goes from a blurry, tilted crop to a validated number.
11 labelled objects, 15 detections. Green is correct, red is a false positive. Move the threshold and watch precision trade against recall.
Python, OpenCV and PyTorch are the base. Filter by area.