Panoptic Image Annotation with a Collaborative Assistant
This work addresses the time-consuming task of image annotation for panoptic segmentation, which is crucial for training computer vision models, and represents an incremental improvement over existing assisted methods.
This paper tackles the problem of reducing annotation time for panoptic segmentation by introducing a collaborative assistant that reacts to annotator actions, achieving speeds 2.4x to 5x faster than manual polygon drawing and outperforming recent machine-assisted interfaces.
This paper aims to reduce the time to annotate images for panoptic segmentation, which requires annotating segmentation masks and class labels for all object instances and stuff regions. We formulate our approach as a collaborative process between an annotator and an automated assistant who take turns to jointly annotate an image using a predefined pool of segments. Actions performed by the annotator serve as a strong contextual signal. The assistant intelligently reacts to this signal by annotating other parts of the image on its own, which reduces the amount of work required by the annotator. We perform thorough experiments on the COCO panoptic dataset, both in simulation and with human annotators. These demonstrate that our approach is significantly faster than the recent machine-assisted interface of [4], and 2.4x to 5x faster than manual polygon drawing. Finally, we show on ADE20k that our method can be used to efficiently annotate new datasets, bootstrapping from a very small amount of annotated data.