Rotation-Invariant Spherical Watermarking via Third-Order SO(3) Representation Coupling
Not on the map. Why?
Abstract (arXiv)
Reliable watermarking of panoramic imagery is fundamentally challenged by arbitrary 3D rotations. As panoramas are defined on the sphere, they naturally transform under the action of SO(3), rendering conventional planar representations and augmentation-based robustness strategies inadequate and devoid of theoretical guarantees. To address this, we formulate panoramas as spherical signals and leverage SO(3) representation theory to derive provably rotation-invariant descriptors. While spherical harmonic coefficients transform equivariantly under rotations, the natural invariant constructions are typically limited to zeroth-order statistics which eliminate directional information and severely constrain embedding capacity. In this work, we introduce a principled third-order invariant construction by coupling higher-order SO(3) irreducible representations via tensor products and projecting onto the trivial representation. This yields a spherical invariant bispectrum that preserves phase information while remaining strictly rotation-invariant. Leveraging this property, we embed watermarks into higher-order spherical harmonic coefficients and recover them from invariant bispectral scalars, enabling reliable extraction under arbitrary 3D rotations. We provide a theoretical proof of SO(3) invariance for it and demonstrate experimentally its near-perfect robustness to continuous rotations while maintaining high visual fidelity.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 27 of its 47 references that are in the map. First posted on arXiv 26 May 2026.
Builds on
27 of 47 references.
References · 27
- 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 735
- 2021E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentialscited by 701
- 2018Spherical CNNscited by 452
- 2018HiDDeN: Hiding Data With Deep Networkscited by 384
- 2019StegaStamp: Invisible Hyperlinks in Physical Photographscited by 256
- 2017Learning SO(3) Equivariant Representations with Spherical CNNscited by 244
- 2017Learning Steerable Filters for Rotation Equivariant CNNscited by 193
- 2022e3nn: Euclidean Neural Networkscited by 171
- 2018Clebsch-Gordan Nets: a Fully Fourier Space Spherical Convolutional Neural Networkcited by 168
- 2023EditGuard: Versatile Image Watermarking for Tamper Localization and Copyright Protectioncited by 109
- 2024Robust Watermarking Using Generative Priors Against Image Editing: From Benchmarking to Advancescited by 105
- 2020Distortion Agnostic Deep Watermarkingcited by 88
15 more
- 2024DreamScene360: Unconstrained Text-to-3D Scene Generation with Panoramic Gaussian Splattingcited by 81
- 2025A Comprehensive Survey on World Models for Embodied AIcited by 65
- 2019Effective Rotation-invariant Point CNN with Spherical Harmonics kernelscited by 57
- 2023TrustMark: Universal Watermarking for Arbitrary Resolution Imagescited by 50
- 2023SepMark: Deep Separable Watermarking for Unified Source Tracing and Deepfake Detectioncited by 43
- 2022Scalable and Equivariant Spherical CNNs by Discrete-Continuous (DISCO) Convolutionscited by 23
- 2024Robust-Wide: Robust Watermarking against Instruction-driven Image Editingcited by 20
- 2024GenEx: Generating an Explorable Worldcited by 20
- 2024GuardSplat: Efficient and Robust Watermarking for 3D Gaussian Splattingcited by 18
- 2025PANORAMA: The Rise of Omnidirectional Vision in the Embodied AI Eracited by 12
- 2025QuaDreamer: Controllable Panoramic Video Generation for Quadruped Robotscited by 7
- 2023A General Framework for Robust G-Invariance in G-Equivariant Networkscited by 5
- 2024The Selective G-Bispectrum and its Inversion: Applications to G-Invariant Networkscited by 5
- 2025Deep Neural Watermarking for Robust Copyright Protection in 3D Point Cloudscited by 3
- 2026Rel-Zero: Harnessing Patch-Pair Invariance for Robust Zero-Watermarking Against AI Editingcited by 1
Built on by
Citing papers per year: 2026: 1.
1 paper.