![[2602.18863] TIACam: Text-Anchored Invariant Feature Learning with Auto-Augmentation for Camera-Robust Zero-Watermarking](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.18863] TIACam: Text-Anchored Invariant Feature Learning with Auto-Augmentation for Camera-Robust Zero-Watermarking
Summary
The paper presents TIACam, a novel framework for camera-robust zero-watermarking that utilizes text-anchored invariant feature learning with auto-augmentation to enhance watermark extraction accuracy.
Why It Matters
As digital watermarking becomes increasingly important for protecting intellectual property, TIACam addresses challenges posed by camera recapture, offering a robust solution that integrates advanced feature learning and semantic consistency, which is crucial for multimedia security.
Key Takeaways
- TIACam integrates a learnable auto-augmentor for camera-like distortions.
- It employs text-anchored invariant feature learning for semantic alignment.
- The framework achieves state-of-the-art stability in feature extraction.
- Zero-watermarking is performed without altering original image pixels.
- Extensive experiments validate TIACam's effectiveness in real-world scenarios.
Electrical Engineering and Systems Science > Image and Video Processing arXiv:2602.18863 (eess) [Submitted on 21 Feb 2026] Title:TIACam: Text-Anchored Invariant Feature Learning with Auto-Augmentation for Camera-Robust Zero-Watermarking Authors:Abdullah All Tanvir, Agnibh Dasgupta, Xin Zhong View a PDF of the paper titled TIACam: Text-Anchored Invariant Feature Learning with Auto-Augmentation for Camera-Robust Zero-Watermarking, by Abdullah All Tanvir and 2 other authors View PDF HTML (experimental) Abstract:Camera recapture introduces complex optical degradations, such as perspective warping, illumination shifts, and Moiré interference, that remain challenging for deep watermarking systems. We present TIACam, a text-anchored invariant feature learning framework with auto-augmentation for camera-robust zero-watermarking. The method integrates three key innovations: (1) a learnable auto-augmentor that discovers camera-like distortions through differentiable geometric, photometric, and Moiré operators; (2) a text-anchored invariant feature learner that enforces semantic consistency via cross-modal adversarial alignment between image and text; and (3) a zero-watermarking head that binds binary messages in the invariant feature space without modifying image pixels. This unified formulation jointly optimizes invariance, semantic alignment, and watermark recoverability. Extensive experiments on both synthetic and real-world camera captures demonstrate that TIACam achieves state-...