Open-World Entity Segmentation

TL;DR

This paper introduces Entity Segmentation, a new image segmentation task that focuses on segmenting all visual entities without class labels, improving segmentation quality and generalization across datasets.

Contribution

It proposes a novel class-agnostic, convolutional center-based architecture for entity segmentation, demonstrating superior performance over class-specific models and easy dataset integration.

Findings

Outperforms panoptic segmentation models in segmentation quality

Models trained on multiple datasets generalize well to unseen domains

Eliminates need for label conflict resolution in dataset merging

Abstract

We introduce a new image segmentation task, called Entity Segmentation (ES), which aims to segment all visual entities (objects and stuffs) in an image without predicting their semantic labels. By removing the need of class label prediction, the models trained for such task can focus more on improving segmentation quality. It has many practical applications such as image manipulation and editing where the quality of segmentation masks is crucial but class labels are less important. We conduct the first-ever study to investigate the feasibility of convolutional center-based representation to segment things and stuffs in a unified manner, and show that such representation fits exceptionally well in the context of ES. More specifically, we propose a CondInst-like fully-convolutional architecture with two novel modules specifically designed to exploit the class-agnostic and non-overlapping requirements of ES. Experiments show that the models designed and trained for ES significantly outperforms popular class-specific panoptic segmentation models in terms of segmentation quality. Moreover, an ES model can be easily trained on a combination of multiple datasets without the need to resolve label conflicts in dataset merging, and the model trained for ES on one or more datasets can generalize very well to other test datasets of unseen domains. The code has been released at https://github.com/dvlab-research/Entity/.