dotdot is a free, open source, web-based image and video annotation tool which is used for labeling data for computer vision algorithms. dotdot supports the primary tasks of supervised machine learning: object detection, image classification, and image segmentation. dotdot allows users to annotate data for each of these cases