CycleSegNet: Object Co-segmentation with Cycle Refinement and Region Correspondence

TL;DR

CycleSegNet introduces a novel co-segmentation framework with cycle refinement and region correspondence modules, significantly improving segmentation accuracy across multiple benchmark datasets.

Contribution

The paper proposes a new co-segmentation network with cycle refinement and region correspondence modules, enhancing information exchange between images.

Findings

Outperforms state-of-the-art on four benchmark datasets

Achieves up to 7.7% improvement in accuracy

Demonstrates effective cycle-based iterative refinement

Abstract

Image co-segmentation is an active computer vision task that aims to segment the common objects from a set of images. Recently, researchers design various learning-based algorithms to undertake the co-segmentation task. The main difficulty in this task is how to effectively transfer information between images to make conditional predictions. In this paper, we present CycleSegNet, a novel framework for the co-segmentation task. Our network design has two key components: a region correspondence module which is the basic operation for exchanging information between local image regions, and a cycle refinement module, which utilizes ConvLSTMs to progressively update image representations and exchange information in a cycle and iterative manner. Extensive experiments demonstrate that our proposed method significantly outperforms the state-of-the-art methods on four popular benchmark datasets -- PASCAL VOC dataset, MSRC dataset, Internet dataset, and iCoseg dataset, by 2.6%, 7.7%, 2.2%, and 2.9%, respectively.