Split GCN: Effective Interactive Annotation for Segmentation of Disconnected Instance

TL;DR

Split-GCN is a novel architecture that improves interactive object segmentation by accurately predicting disconnected components using self-attention and topology transformation, reducing annotation costs and enhancing generalization across datasets.

Contribution

It introduces Split-GCN, a new polygon-based model with self-attention that effectively predicts disconnected object components and generalizes well across diverse datasets.

Findings

Achieves competitive performance on Cityscapes.

Outperforms baseline models in disconnected component prediction.

Demonstrates strong cross-domain generalization.

Abstract

Annotating object boundaries by humans demands high costs. Recently, polygon-based annotation methods with human interaction have shown successful performance. However, given the connected vertex topology, these methods exhibit difficulty predicting the disconnected components in an object. This paper introduces Split-GCN, a novel architecture based on the polygon approach and self-attention mechanism. By offering the direction information, Split-GCN enables the polygon vertices to move more precisely to the object boundary. Our model successfully predicts disconnected components of an object by transforming the initial topology using the context exchange about the dependencies of vertices. Split-GCN demonstrates competitive performance with the state-of-the-art models on Cityscapes and even higher performance with the baseline models. On four cross-domain datasets, we confirm our model's generalization ability.