I came across this awesome library of diatom images which would be super helpful for a project where I need to count thousands of diatoms from microscope imagery: Diatoms Dataset Object Detection Dataset and Pre-Trained Model by Diatoms Dataset
I have been trying the pre-trained model in the browser and locally (Mac OS Ventura (M1)) and it has been getting some promising results. However, the available model was built on an old version of the dataset (version 2) which had just a couple hundred images. Now the dataset has over 5000 images and I’d love to get an updated model.
I am very much new to image recognition and AI. I have some background in python for geospatial science but I am a professor of geography and not computer science so I am finding myself lost on a lot of the terminology and best steps to take.
Is anyone able to guide me on how to update the model, or would have sufficient time to update the model online…?