Discrete State Detection

AVEVA Vision AI Assistant uses Supervised Deep Learning Models to distinguish between two states. You provide image datasets containing examples which are already labelled or categorized, and this enables the algorithm to build a general model of each category. The algorithm then processes the real (uncategorized) data and attempts to put each item into one of the pre-learned categories.

To derive the best possible categorization, the dataset is trained multiple times. Each training round is referred to as an epoch. You can constantly refine the model to correctly categorize images that are predicted incorrectly. For example, is the toothbrush red or blue?