DA-Mamba: Learning Domain-Aware State Space Model for Global-Local Alignment in Domain Adaptive Object Detection

2026-03-19

AI Summary

This work addresses the limitations of existing domain adaptive object detection methods, which predominantly rely on local feature alignment and struggle to capture global domain-invariant representations, while Transformer-based approaches suffer from high computational complexity that hinders deployment. To overcome these challenges, we propose DA-Mamba, the first framework to introduce State Space Models (SSMs) into domain adaptive detection. Specifically, we design Image-Aware SSM and Object-Aware SSM modules to enable joint global-local alignment at the image level in the backbone and at the instance level in the detection head, respectively. Our approach achieves linear computational complexity while significantly improving cross-domain detection performance, demonstrating efficient and robust domain adaptation across multiple benchmarks.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Computer Vision: Object Detection & Categorization
• Application Domains: Security

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Domain Adaptive Object Detection (DAOD) aims to transfer detectors from a labeled source domain to an unlabeled target domain. Existing DAOD methods employ multi-granularity feature alignment to learn domain-invariant representations. However, the local connectivity of their CNN-based backbone and detection head restricts alignment to local regions, failing to extract global domain-invariant features. Although transformer-based DAOD methods capture global dependencies via attention mechanisms, their quadratic computational cost hinders practical deployment. To solve this, we propose DA-Mamba, a hybrid CNN-State Space Models (SSMs) architecture that combines the efficiency of CNNs with the linear-time long-range modeling capability of State Space Models (SSMs) to capture both global and local domain-invariant features. Specifically, we introduce two novel modules: Image-Aware SSM (IA-SSM) and Object-Aware SSM (OA-SSM). IA-SSM is integrated into the backbone to enhance global domain awareness, enabling image-level global and local alignment. OA-SSM is inserted into the detection head to model spatial and semantic dependencies among objects, enhancing instance-level alignment. Comprehensive experiments demonstrate that the proposed method can efficiently improve the cross-domain performance of the object detector.

Problem

• Domain Adaptive Object Detection
• Global-Local Alignment
• Domain-Invariant Features
• Computational Efficiency
• State Space Models

Innovation

• State Space Model
• Domain Adaptive Object Detection
• Global-Local Alignment
• Linear-Time Long-Range Modeling
• Hybrid CNN-SSM Architecture