Ri-mamba: Rotation-invariant Mamba For Robust Text-to-shape Retrieval

Abstract

3D assets have rapidly expanded in quantity and diversity due to the growing popularity of virtual reality and gaming. As a result, text-to-shape retrieval has become essential in facilitating intuitive search within large repositories. However, existing methods require canonical poses and support few object categories, limiting their real-world applicability where objects can belong to diverse classes and appear in random orientations. To address this challenge, we propose RI-Mamba, the first rotation-invariant state-space model for point clouds. RI-Mamba defines global and local reference frames to disentangle pose from geometry and uses Hilbert sorting to construct token sequences with meaningful geometric structure while maintaining rotation invariance. We further introduce a novel strategy to compute orientational embeddings and reintegrate them via feature-wise linear modulation, effectively recovering spatial context and enhancing model expressiveness. Our strategy is inherently

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keynguyen2026ri

Related papers

• Risa-net: Rotation-invariant Structure-aware Network For Fine-grained 3D Shape Retrieval (2020)5.48
• Riemann-based Multi-scale Attention Reasoning Network For Text-3d Retrieval (2024)4.53
• Hyperbolic Hierarchical Alignment Reasoning Network For Text-3d Retrieval (2025)1.81
• SAMURAI: Shape-aware Multimodal Retrieval For 3D Object Identification (2025)0.00
• Enhanced Cross-modal 3D Retrieval Via Tri-modal Reconstruction (2025)0.00
• MUSE: Mamba Is Efficient Multi-scale Learner For Text-video Retrieval (2024)6.34
• Mambahash: Visual State Space Deep Hashing Model For Large-scale Image Retrieval (2025)3.95
• Parts2words: Learning Joint Embedding Of Point Clouds And Texts By Bidirectional Matching Between Parts And Words (2021)9.96