Hua Sun

Large language models (LLMs) pose significant challenges to content authentication, as their sophisticated generation capabilities make distinguishing AI-produced text from human writing increasingly difficult. Current detection methods suffer from limited information capture, poor rate–distortion trade-offs, and vulnerability to adversarial perturbations. We present CurveMark, a novel dual-channel detection framework that combines probability curvature analysis with dynamic semantic watermarking, grounded in information-theoretic principles to maximize mutual information between text sources and observable features. To address the limitation of requiring prior knowledge of source models, we incorporate a Bayesian multi-hypothesis detection framework for statistical inference without prior assumptions. Our approach embeds imperceptible watermarks during generation via entropy-aware, semantically informed token selection and extracts complementary features from probability curvature patterns and watermark-specific metrics. Evaluation across multiple datasets and LLM architectures demonstrates 95.4% detection accuracy with minimal quality degradation (perplexity increase < 1.3), achieving 85–89% channel capacity utilization and robust performance under adversarial perturbations (72–94% information retention).

Identifying template watermarks under severe geometric distortions is a significant scientific problem in the current watermarking research for remote sensing images. We propose a novel watermarking algorithm that integrates the ring-shaped template watermark with the multiscale local contrast measure (LCM) method. In the embedding stage, the ring-shaped template watermark is embedded into the discrete Fourier transform (DFT) magnitude coefficients, converting the watermark into small targets in the DFT domain. During the detection stage, the multiscale LCM, a classic infrared small target detection method, enhances these small targets and generates a contrast map. Peak detection is then performed on the contrast map to determine the radius of the template watermark. Finally, circular edge local binarization is applied to extract the watermark information. The proposed method enables synchronization recovery of watermarks under blind conditions. The experimental results demonstrate that the method possesses strong robustness against various geometric attacks such as rotation, scaling, translation, and cropping. It outperforms comparative algorithms in terms of robustness and also exhibits good imperceptibility.

In order to safeguard image copyrights, zero-watermarking technology extracts robust features and generates watermarks without altering the original image. Traditional zero-watermarking methods rely on handcrafted feature descriptors to enhance their performance. With the advancement of deep learning, this paper introduces “ZWNet”, an end-to-end zero-watermarking scheme that obviates the necessity for specialized knowledge in image features and is exclusively composed of artificial neural networks. The architecture of ZWNet synergistically incorporates ConvNeXt and LK-PAN to augment the extraction of local features while accounting for the global context. A key aspect of ZWNet is its watermark block, as the network head part, which fulfills functions such as feature optimization, identifier output, encryption, and copyright fusion. The training strategy addresses the challenge of simultaneously enhancing robustness and discriminability by producing the same identifier for attacked images and distinct identifiers for different images. Experimental validation of ZWNet’s performance has been conducted, demonstrating its robustness with the normalized coefficient of the zero-watermark consistently exceeding 0.97 against rotation, noise, crop, and blur attacks. Regarding discriminability, the Hamming distance of the generated watermarks exceeds 88 for images with the same copyright but different content. Furthermore, the efficiency of watermark generation is affirmed, with an average processing time of 96 ms. These experimental results substantiate the superiority of the proposed scheme over existing zero-watermarking methods.