# JND-Guided Neural Watermarking with Spatial Transformer Decoding for Screen-Capture Robustness

Source: [JND-Guided Neural Watermarking with Spatial Transformer Decoding for Screen-Capture Robustness](<https://www.sota2.com/research/paper/69cb551c0636e5bf4afc9f33>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-01T05:00:04.436Z

Authors: Jiayi Qin, Jingwei Li, Chuan Wu

Published: 2026-03-24T03:38:42.000Z

## Abstract

Screen-shooting robust watermarking aims to imperceptibly embed extractable information into host images such that the watermark survives the complex distortion pipeline of screen display and camera recapture. However, achieving high extraction accuracy while maintaining satisfactory visual quality remains an open challenge, primarily because the screen-shooting channel introduces severe and entangled degradations including Moir\\'{e} patterns, color-gamut shifts, perspective warping, and sensor noise. In this paper, we present an end-to-end deep learning framework that jointly optimizes watermark embedding and extraction for screen-shooting robustness. Our framework incorporates three key innovations: (i) a comprehensive noise simulation layer that faithfully models realistic screen-shooting distortions -- notably including a physically-motivated Moir\\'{e} pattern generator -- enabling the network to learn robust representations against the full spectrum of capture-channel noise through adversarial training; (ii) a Just Noticeable Distortion (JND) perceptual loss function that adaptively modulates watermark embedding strength by supervising the perceptual discrepancy between the JND coefficient map and the watermark residual, thereby concentrating watermark energy in perceptually insensitive regions to maximize visual quality; and (iii) two complementary automatic localization modules -- a semantic-segmentation-based foreground extractor for captured image rectification and a symmetric noise template mechanism for anti-cropping region recovery -- that enable fully automated watermark decoding under realistic deployment conditions. Extensive experiments demonstrate that our method achieves an average PSNR of 30.94~dB and SSIM of 0.94 on watermarked images while embedding 127-bit payloads.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Watermarking Robustness on Robustness (test)](<https://www.sota2.com/research/sota/watermarking-robustness-on-robustness-test.md>): Watermarking Robustness; Robustness (test); BER (40–49 cm): 2.55.

- [Watermarking Robustness on Unspecified robustness evaluation set (test)](<https://www.sota2.com/research/sota/watermarking-robustness-on-unspecified-robustness-evaluation-set-test.md>): Watermarking Robustness; Unspecified robustness evaluation set (test); BER (−40–55°): 1.59.

- [Watermarking Robustness on robustness evaluation set (test)](<https://www.sota2.com/research/sota/watermarking-robustness-on-robustness-evaluation-set-test.md>): Watermarking Robustness; robustness evaluation set (test); BER (0.05 Ratio): 1.89.

- [Image Watermarking on Watermarked Images](<https://www.sota2.com/research/sota/image-watermarking-on-watermarked-images.md>): Image Watermarking; Watermarked Images; PSNR: 31.77.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.26766>)

[PDF](<https://arxiv.org/pdf/2603.26766.pdf>)
