AutoML Vision provides insights from images at the edge and cloud. Pre-trained Vision API models can also be used to understand text and detect emotion. Google Cloud offers two computer vision products, which use machine learning to help understand your images with an industry-leading prediction accuracy. Automate the creation of custom machine learning models. Upload images, train custom image models using AutoML Vision's intuitive graphical interface, optimize your models for accuracy and latency, and export them to your cloud application or to a range of devices at the edge. Google Cloud's Vision API provides powerful pre-trained machine-learning models via REST and RPC APIs. Assign labels to images and classify them quickly into millions of predefined groups. Detect faces and objects, read printed and handwritten texts, and add valuable metadata to your image catalog.