WIMFRIS: WIndow Mamba Fusion and Parameter Efficient Tuning for Referring Image Segmentation

Existing Parameter-Efficient Tuning (PET) methods for Referring Image Segmentation (RIS) primarily focus on layer-wise feature alignment, often neglecting the crucial role of a neck module for the intermediate fusion of aggregated multi-scale features, which creates a significant performance bottleneck. To address this limitation, we introduce WIMFRIS, a novel framework that establishes a powerful neck architecture alongside a simple yet effective PET strategy. At its core is our proposed Hierarchical Mamba Fusion (HMF) block, which first aggregates multi-scale features and then employs a novel Window Mamba Fuser (WMF) module to perform effective intermediate fusion. This WMF module leverages non-overlapping window partitioning to mitigate the information decay problem inherent in State-Space Models (SSMs) while ensuring rich local-global context interaction. Furthermore, our PET strategy enhances primary alignment with a Mamba Text Adapter (MTA) for robust textual priors, a Multi-Scale Aligner (MSA) for precise vision-language fusion, and learnable emphasis parameters for adaptive stage-wise feature weighting. Extensive experiments demonstrate that WIMFRIS achieves new state-of-the-art performance across all public RIS benchmarks. Code is available here.