H2O Vision Services: Label Genie to Document AI

This lesson walks through H2O's vision services, starting with Label Genie clustering a set of car and coffee images to accelerate annotation before the clusters are exported for training elsewhere. It moves to H2O Hydrogen Torch, reviewing a segmentation experiment's training charts, validation predictions, and interpretation insights that highlight which coin features drove predictions, before deploying the model to a live scoring endpoint. The lesson closes with H2O Document AI, walking through OCR, labeling regions such as dates of birth and phone numbers, training a token labeling model, and publishing it to a scoring pipeline.