700 - Reducing the Annotation Effort for Video Object Segmentation Datasets