ReMamber: Referring Image Segmentation with Mamba Twister

TL;DR

ReMamber introduces an efficient, linear-complexity architecture for referring image segmentation by integrating a novel Mamba Twister block that explicitly models image-text interactions and fuses multi-modal features.

Contribution

The paper proposes ReMamber, a new RIS model that combines Mamba with a multi-modal Twister block for effective and resource-efficient image-text feature fusion.

Findings

Achieves competitive results on three benchmarks.

Demonstrates the effectiveness of the Mamba Twister for multi-modal fusion.

Provides analysis and discussion on fusion design options.

Abstract

Referring Image Segmentation~(RIS) leveraging transformers has achieved great success on the interpretation of complex visual-language tasks. However, the quadratic computation cost makes it resource-consuming in capturing long-range visual-language dependencies. Fortunately, Mamba addresses this with efficient linear complexity in processing. However, directly applying Mamba to multi-modal interactions presents challenges, primarily due to inadequate channel interactions for the effective fusion of multi-modal data. In this paper, we propose ReMamber, a novel RIS architecture that integrates the power of Mamba with a multi-modal Mamba Twister block. The Mamba Twister explicitly models image-text interaction, and fuses textual and visual features through its unique channel and spatial twisting mechanism. We achieve competitive results on three challenging benchmarks with a simple and efficient architecture. Moreover, we conduct thorough analyses of ReMamber and discuss other fusion designs using Mamba. These provide valuable perspectives for future research. The code has been released at: https://github.com/yyh-rain-song/ReMamber.