Towards Stable Co-saliency Detection and Object Co-segmentation

TL;DR

This paper introduces a stable recurrent model with a novel loss function for improved co-saliency detection and object co-segmentation, addressing inter-image relation modeling and order sensitivity issues.

Contribution

The paper proposes a multi-path stable recurrent unit (MSRU) with dummy orders mechanisms and a cross-order contrastive loss (COCL) to enhance stability and robustness in co-saliency detection and co-segmentation.

Findings

Outperforms state-of-the-art methods on five CoSOD datasets.

Demonstrates improved stability and robustness in inference.

Effective in reducing order sensitivity in inter-image relation modeling.

Abstract

In this paper, we present a novel model for simultaneous stable co-saliency detection (CoSOD) and object co-segmentation (CoSEG). To detect co-saliency (segmentation) accurately, the core problem is to well model inter-image relations between an image group. Some methods design sophisticated modules, such as recurrent neural network (RNN), to address this problem. However, order-sensitive problem is the major drawback of RNN, which heavily affects the stability of proposed CoSOD (CoSEG) model. In this paper, inspired by RNN-based model, we first propose a multi-path stable recurrent unit (MSRU), containing dummy orders mechanisms (DOM) and recurrent unit (RU). Our proposed MSRU not only helps CoSOD (CoSEG) model captures robust inter-image relations, but also reduces order-sensitivity, resulting in a more stable inference and training process. { Moreover, we design a cross-order contrastive loss (COCL) that can further address order-sensitive problem by pulling close the feature embedding generated from different input orders.} We validate our model on five widely used CoSOD datasets (CoCA, CoSOD3k, Cosal2015, iCoseg and MSRC), and three widely used datasets (Internet, iCoseg and PASCAL-VOC) for object co-segmentation, the performance demonstrates the superiority of the proposed approach as compared to the state-of-the-art (SOTA) methods.