Approach to marking images

In this paper propose the method to speed up labeling of the image for creating new datasets. We need datasets for training convolutional neural networks that solve the various task of computer vision for example: image classification, object detection, and localization..