Fast, Secure, And High-Capacity Image Watermarking With Text Autoencoded Text Vectors

ICLR 2026 Conference Submission9389 Authors

Keywords: Computer Vision, Watermarking, Autoencoder, Security, Representation Learning

TL;DR: LatentSeal turns watermarking into semantic communication: encode a sentence into a 256-D unit-norm latent, secure it via key-conditioned rotation, embed robustly to survive common edits, decode in real time with calibrated AUROC 0.97–0.99.

Abstract:

Most image watermarking systems focus on robustness, capacity, and imperceptibility while treating the embedded payload as meaningless bits. This bit-centric view imposes a hard ceiling on capacity and prevents watermarks from carrying useful information. We propose LatentSeal, which reframes watermarking as semantic communication: a lightweight text autoencoder maps full-sentence messages into a compact 256-dimensional unit-norm latent vector, which is robustly embedded by a finetuned watermark model and secured through a secret, invertible rotation.
The resulting system hides full-sentence messages, decodes in real time, and survives valuemetric and geometric attacks. It surpasses prior state of the art in BLEU-4 and Exact Match on several benchmarks, while breaking through the long-standing 256-bit empirical payload ceiling. It also introduces a statistically calibrated score that yields a ROC AUC score of 0.97-0.99, and practical operating points for deployment.
By shifting from bit payloads to semantic latent vectors, LatentSeal enables watermarking that is not only robust and high-capacity, but also secure and interpretable, providing a concrete path toward provenance, tamper explanation, and trustworthy AI governance. Models, training and inference code, and data splits will be available upon publication (code available in a the supplementary zip file).

Primary Area: applications to computer vision, audio, language, and other modalities

Submission Number: 9389