Overview

Classification model can be integrated into labeling interface and significantly speed up manual classification / tagging of images and objects:

1. Model predicts most probable tags
2. Labeler visually compare predictions with target image/object and assign the correct tag
3. Also app allows to review and correct already assigned tags

This approach significantly outperforms manual labeling especially when annotator has to select one of hundred or even thousand of tags. For example, in retail there are huge catalogs, and it is time-consuming and almost impossible to tag objects on shelves without special tool.

How To Run

1. Add app from Ecosystem to your team if it is not there
2. Deploy classification model
3. Run app from labeling interface

How To Use