188 results about "Watermark method" patented technology

The invention relates to a method for generating an image watermark based on potential space multi-scale feature modulation. The method mainly comprises a watermark residual embedding and extracting method based on potential space multi-scale feature modulation, a watermark intensity regulation and control strategy based on time sequence gating and a robust watermark end-to-end training framework based on potential space. According to the method, watermark information is fused by using a multi-scale structure in a potential space of a diffusion model, meanwhile, a time sequence gating strategy dynamically determines the embedding time step of a watermark by calculating the convergence degree of a potential vector, and the image quality and the watermark robustness are balanced. Sigmoid weighted noise prediction loss is introduced in the model training process, and denoising control over low time steps is enhanced. The method can be widely applied to scenes such as image generation, copyright identification and content traceability, and has good practicability and popularization prospects.

The invention relates to the field of watermarking algorithms, and provides a watermarking method and system for generating a text by a large language model. The method comprises the steps that a prompt text and a generated text sequence are input into a generation model, and an original logic score of a next mark position is generated through probability distribution calculation; performing semantic feature extraction on the generated text sequence by using an embedding model to obtain semantic embedding representation; converting the semantic embedding representation into a watermark logic score based on a pre-trained watermark model; and performing weighted combination on the original logic score and the watermark logic score to generate a final logic score, and outputting a watermark mark based on the final logic score. According to the method, the high robustness of the watermark is improved, and the safety of the watermark is also improved.

The invention discloses a diffusion model robust watermarking method based on affine coupling flow accurate coding. The method comprises the following steps: S1, training a watermark encoder and a watermark decoder; s2, freezing the trained watermark encoder and the diffusion model to perform diffusion iteration on the initial hidden variable to obtain a watermark-containing image, introducing an attack layer to finely adjust the pre-trained watermark decoder to generate the watermark-containing image, performing reverse diffusion of the diffusion model to form the initial hidden variable, and extracting watermark information by the pre-trained watermark decoder; and S3, uniformly and integrally deploying the trained watermark encoder and the diffusion model, and extracting the watermark in the user generated image to verify the identity. According to the diffusion model robust watermarking method based on affine coupling flow accurate coding, the problems that the robustness of a reconstruction watermark removing technology is insufficient, and the presentation of watermark information in an image generated by a user is single are solved.

A computer-implemented method for watermarking images includes providing a secret key network (SKN) that is adapted to output a standard multivariate normal (SMVN) distribution for a given input image distribution, applying an input image to the SKN, generating a secret key signature (SKS) as a real vector, and embedding a watermark in the input image by using an adversarial attack to modify the input image in a manner that aligns the SKN's output with the SKS. A computer-implemented method for detecting a watermark in an image is also provided.

The invention discloses an image watermarking method based on a space channel interactive attention mechanism. The image watermarking method mainly comprises the following steps: acquiring an original image and constructing an image watermarking network based on the space channel attention mechanism; inputting the original image and the watermark into a watermark-containing image generation framework to obtain a watermark-containing image; inputting the image with the watermark into an image processing noise layer to obtain a noise image; inputting the noise image into a watermark extraction framework to obtain an extracted watermark; and constraining the training of the whole image watermark network based on the space channel attention mechanism by using the total loss. The image watermark network based on the space channel attention mechanism is used for copyright protection, and high-capacity embedding of watermarks, high-quality generation of images with watermarks and high-robustness extraction of the watermarks are achieved. According to the method, a depth feature map of the watermark is generated through watermark dimension extension, remodeling and diffusion by using a multi-layer sensor and the like. Meanwhile, the feature retention loss can promote the optimization of the total loss on the network training.

The invention discloses a robust reversible watermarking method based on BFMs and adaptive normalization. The method comprises the steps of watermark embedding and watermark extraction. The watermark embedding specifically comprises the steps of generating watermark sequence information, performing integer wavelet transform on a carrier image, extracting a low-frequency region, and calculating BFMs of the low-frequency region; a proper watermark pair is selected through a private key, self-adaptive normalization is carried out according to the watermark pair, and the watermark is embedded in the normalized amplitude; reconstructing an image based on the amplitude of the embedded watermark; constructing a final reversible watermark image based on the reconstructed image; watermark extraction: judging whether the image is attacked or not; for an unattacked image, carrying out watermark extraction based on the auxiliary information; for an attacked image, watermark extraction is carried out by calculating the BFMs of the attacked image. The method has noise resistance and also has relatively good image reconstruction capability.

The invention provides an anti-screen-shooting robust text image watermarking method fusing edge attention gating and multi-scale cavity convolution, which comprises the following steps: step 1, designing a message processor for preprocessing a secret message, adding an expansion network ExpandNet and a channel attention SENet in the message processor, and carrying out edge attention gating and multi-scale cavity convolution on the expansion network ExpandNet and the channel attention SENet; inputting a secret message in a binary form and outputting a message feature map; step 2, constructing a text watermark framework based on edge attention gating and multi-scale cavity convolution fusion; adding a functional module in an encoder of the end-to-end watermark framework; 3, designing a noise layer, and performing simulation attack on the model in the watermark model training process; and step 4, designing an optimization mode of the loss function. According to the method, the robustness and imperceptibility of the watermark are remarkably improved.

The invention discloses a vision enhancement robust digital dark watermarking method based on deep learning, a storage medium and equipment, and belongs to the technical field of digital image processing and information security. According to the method, a visual enhancement module comprising a self-adaptive watermark region adjustment module and a frequency enhancement module is constructed, and a multi-region local loss training mechanism and a screen shooting simulation module are combined, so that the visual quality of a watermark image is improved, and meanwhile, the robustness of the watermark image for resisting physical attacks such as screen shooting is enhanced. The specific processing flow comprises the following steps: acquiring an original image and watermark information; embedding the watermark information into the image by using an encoder, wherein an embedded region is optimized by a self-adaptive region guide matrix generated based on the image edge and gray information; in the model training process, a noise layer simulating screen shooting physical distortion is introduced, and the area weight is dynamically adjusted according to residual error distribution in the later stage of training so as to focus and optimize the image quality; watermark information is extracted from an image which may suffer from an attack by using a decoder.

The invention provides a high-fidelity anti-compression image watermarking method and system based on spectrum-airspace decoupling, and belongs to the field of information security. Firstly, the watermark information is mapped and remodeled; a watermark encoder based on multi-granularity spectrum-spatial domain feature decoupling is constructed, discrete cosine transform is introduced to filter out high-frequency components, Haar wavelet transform is adopted to realize lossless downsampling, a fast Fourier transform dynamic filter is combined to capture global semantic features, and local texture details are combined through multi-scale spatial domain volume accumulation; designing a physical perception and visual self-adaptive dual embedding strategy, and anchoring watermark energy to an anti-compression brightness channel; constructing an anti-attack layer containing differentiable JPEG compression simulation and mixed noise simulation, and participating in network training; and constructing a decoder and designing a loss function to carry out network optimization. According to the method, the robustness of the watermark under strong compression and complex black box attacks is improved, and extremely high visual imperceptibility is realized through physical and visual constraints.

The invention discloses a high-fidelity robust watermarking method based on deep reinforcement learning. The method is composed of a reinforcement learning module, a watermark encoding and decoding module and a watermark synchronization module in sequence. The reinforcement learning module automatically selects an optimal watermark embedding position in a local area of the image through an adaptive decision-making mechanism of a deep Q network to realize dynamic optimization of an embedding strategy; the watermark encoding and decoding module adopts a convolutional block attention mechanism (CBAM) and a multi-scale feature fusion structure, and realizes high-fidelity embedding and accurate extraction of watermark information while maintaining the visual quality of the image; the watermark synchronization module is based on a U2Net model structure, and completes spatial alignment of a watermark area through mask prediction and affine correction, thereby effectively solving positioning deviation caused by geometric attacks such as rotation, zooming, cutting and the like. The three components work cooperatively, and high watermark extraction precision can be kept under various compression, noise and geometric disturbance conditions. The method has the advantages of imperceptible embedding, high extraction stability, independent training and combined deployment of modules and the like, and can be widely applied to the fields of digital image copyright protection, AIGC (Artificial Intelligence Generation Content) traceability, multimedia authenticity verification and the like.

To discriminate a generation source of text data even after a digital watermark key has leaked.SOLUTION: In a double digital watermark system 1, a digital watermark embedding device classifies, by a first key and a hash function, a plurality of tokens generated next from a first token into a first group or a second group; receives tokens of the first group; classifies, by a second key and the hash function, a plurality of tokens generated next from a second token of a token sequence into a third group or a fourth group; and increases probabilities of tokens of the first and third groups. A digital watermark detector calculates, by the first token and the first key, the number of tokens of the first group; calculates, by the second key, the number of tokens of the third group; and determines presence or absence of first and second digital watermarks on the basis of a first score obtained by a statistical test from the calculated numbers of tokens and a total number of tokens, and a second score obtained by a statistical test from the number of tokens of the first group.SELECTED DRAWING: Figure 1

The invention discloses a text watermarking method and device, a storage medium and electronic equipment, and relates to the technical field of artificial intelligence and digital watermarking, and the method comprises the steps: obtaining an original text and a watermark, inputting the original text and the watermark into an embedded watermark model, and carrying out the watermark embedding operation to obtain a watermark text. The watermark embedding model comprises a gating modification mechanism, the gating modification mechanism is used for replacing a segmented word with a threshold value greater than a preset threshold value in the original text with a candidate segmented word with the same meaning as the segmented word, the watermark is embedded in the candidate segmented word, and the threshold value is used for representing the feasibility of embedding the watermark at the position of the segmented word. According to the technical scheme, through the gating modification mechanism and the candidate word segmentation generation strategy, the text modification degree is strictly controlled, the semantic coherence of the text embedded with the watermark is improved, and the watermark robustness is improved.

The invention discloses an anti-multi-attack medical image robust watermarking method based on a Mama architecture, which comprises the following steps: an original medical image and watermark information are processed through an encoder network to generate a watermark-containing image, and an encoder comprises a hierarchical feature refined sampler module and a channel perception Mama architecture bottleneck layer; a watermark is recovered from a watermark-containing image which may be attacked through a decoder network, and a decoder carries out dual-path processing through a wavelet domain Mama module and a sequence context dependence module; randomly extracting 2-4 attacks from the attack library to apply dynamic composite attacks to the watermark-containing image in a random sequence and intensity; and performing end-to-end optimization through a composite loss function including reconstruction loss, clean decoding loss, robust decoding loss and perception loss. According to the method, the image frequency band is decoupled through discrete wavelet transform, meanwhile, the watermark is embedded at the high frequency, the long-range dependency relationship of the wavelet coefficient is modeled by using the selective state space mechanism of Mamba, and the robustness of the watermark is improved.

The invention relates to a robust watermarking method of a diffusion model based on an error correction code, which belongs to the technical field of large model watermarking, and comprises the following steps of: encoding a watermark message by using a Reed-Solomon error correction code to generate a code word containing redundancy check information; carrying out randomization processing on the code word to generate a pseudo-random bit stream; embedding the pseudo-random bit stream into a variable of an initial Gaussian potential space of a diffusion model by adopting a distribution preserving sampling technology; and finally, generating an image with a watermark through a denoising process of the diffusion model. When the watermark is extracted, reverse diffusion processing is carried out on an image which is possibly damaged to recover an approximate value of a potential space variable, a code word which is possibly damaged is extracted from the approximate value, and original watermark information is recovered by utilizing the error correction capability of a Reed-Solomon decoder. On the premise of keeping the generation performance, the robustness of the watermark to compression, noise, cutting and other potential tampering attacks is enhanced.

The invention discloses an image watermarking method for a stable diffusion model, and the method comprises the steps: selecting one of all potential features in a reverse diffusion process of the stable diffusion model as a target potential feature, and constructing an image watermark generation recognition model which comprises a watermark information embedding module and a watermark extraction module; the watermark information embedding module is used for embedding user watermark information into the target potential features, and inputting the obtained watermark target potential features instead of the original target potential features into the remaining sub-models of the stable diffusion model to generate a watermark image; and the watermark information extraction module is used for extracting user watermark information from the watermark potential features obtained by carrying out DDIM inversion on the watermark image, and the user generating the watermark image can be identified according to the extracted user watermark information. According to the method, the user watermark information is embedded into the potential features of the stable diffusion model, so that the image traceability with high robustness and high concealment is realized on the premise of not damaging the quality of the generated image.

The invention discloses a generative adversarial network (GAN)-based recoverable watermarking resisting method, which not only can ensure that an image is protected by privacy and copyright, but also can only allow an authorized party to safely use. Specifically, an encoder-decoder-restorer architecture based on a GAN (Generic Area Network) is innovatively designed, and watermarking-resistant embedding and extraction and image restoration are established as a unified task. And a dual optimization strategy is designed to resist loss and a dynamic joint training strategy, so as to achieve ideal balance among image quality, copyright protection, privacy protection and image recovery capability. Experimental results show that the method provided by the invention can ensure that the authorized DNN classifier recovers the protected image to perform accurate classification and identification when needed, and meanwhile, excellent privacy and copyright protection capability is also provided.

The invention discloses an image watermarking method and system based on decoding guided by a self-supervised visual model. The method comprises the following steps: acquiring watermark information m and an original image I of each data sample in a training data set; for each data sample, preprocessing the watermark information m to generate a message tensor M, and performing degradation distortion processing on the original image I to generate a distortion guide heat map DGH, thereby obtaining a preprocessed training data set; and performing self-supervising training on the self-supervising visual model by using the preprocessed training data set to finally obtain the trained self-supervising visual model. The invention aims to solve the problem that the decoding robustness is poor when the mask prediction is inaccurate due to the fact that the existing printing-resistant shooting watermark technology excessively depends on the positioning mask, and the semantic comprehension capability of the depth vision model is utilized to replace the traditional geometric synchronization, so that the end-to-end robust decoding without correction is realized.

The invention discloses a robust watermarking method for stable diffusion generated images. The robust watermarking method comprises the following steps: step 1, watermark embedding; step 2, generating a watermark image; step 3, watermark extraction; step 4, watermark detection and source tracing; according to the method, the watermark information is encrypted and mapped into Gaussian noise approximately conforming to normal distribution, and is embedded into a potential space instead of directly modifying pixels, so that visual distortion of an image is avoided, native logic of a diffusion model is fit, model parameters do not need to be adjusted or network fine adjustment is not needed, and the original generation performance of the model can be fully guaranteed; the watermark is expanded to the potential space full dimension, the watermark covers the whole semantic level of the image, a voting mechanism is introduced in the extraction stage to carry out consistency verification on the multi-copy watermark, various common attacks can be effectively resisted, the generated image can still keep the complete content structure and semantic consistency even under the high watermark capacity, and the image quality is improved. And the balance of model performance, image quality and robustness is realized.

The invention relates to an implicit black box watermarking method for copyright protection of a multi-modal model, and belongs to the technical field of artificial intelligence safety. The method specifically comprises the following steps: 1, generating a multi-modal model black box backdoor trigger; 2, completing model training of a conversion module according to the backdoor trigger set in the step 1; and 3, performing ownership verification and infringement model judgment on the suspicious model according to the backdoor trigger set in the step 1 and the conversion module in the step 2. According to the method, the resistance trigger is generated by adopting the original samples in the original data set, the limit of an existing distributed external trigger is broken, the method can adapt to a real application scene of a multi-modal model under the condition of small overhead, and the method has good practicability and expandability. In addition, two-stage watermark verification based on a conversion module is adopted, so that the infringement problem caused by hostile attacks can be effectively avoided.

The invention discloses a depth classification model black box watermarking method for natural deformation of an image bound with an identity. The method comprises the steps of distortion field generation based on an ID, watermark embedding and watermark extraction verification. The generation of the distortion field based on the ID is that a hash value is generated from the ID representing the identity of a model owner, and a unique distortion parameter is associated, so that the distortion field bound with the ID of the model owner is constructed; the watermark embedding comprises an original data set acquisition process, a training data set generation process and a training process, the training data set comprises a clean sample, a trigger sample and a potential attacker sample which are obtained through the action of a distorted field, and the training data set is used for training a depth classification model to obtain a watermark model; in the watermark extraction verification, a test trigger sample is generated by using a distorted field corresponding to a model owner ID, and is input into a watermark model for verification, and watermark embedding has good uniqueness and concealment.

The invention discloses a watermarking method and device suitable for any resolution image input, and the method comprises the steps: generating a fixed low-resolution watermark residual error through a watermark encoder according to an input carrier image and a watermark message, restoring the watermark residual error to the size of the carrier image, and superposing the watermark residual error with the carrier image to obtain a watermark-containing image; designing a watermark discriminator based on the residual neural network, wherein the watermark discriminator is used for distinguishing the difference between the watermark-containing image and the carrier image; a watermark noise layer is used for simulating various distortion operations suffered by an image under a digital channel, and disturbance is applied to the watermark-containing image processed by a watermark discriminator; and transmitting the watermark-containing image to which the disturbance is applied in a real channel, inputting the watermark-containing image into a robust watermark decoder, and decoding and extracting a watermark message. According to the method and the device, image input of any resolution can be processed, good invisibility and robustness can be kept, and copyright protection and leakage traceability of the image of any resolution can be realized.

The invention discloses an image watermark generation method and system based on a diffusion model optimization potential variable, and the method comprises the steps: training a secret encoder and a watermark extractor, and coding a message into a secret residual error; after the residual error is injected into the potential representation of the image, mapping the residual error into a potential vector with a watermark, which accords with the prior of a diffusion model, through DDIM inversion; performing iterative optimization on the potential vector by using a pre-trained diffusion model U-Net, introducing double constraints of image reconstruction loss and message fidelity loss in a denoising process, and generating an image with high visual quality and high watermark robustness; and finally, extracting the message through a watermark extractor and carrying out statistical significance verification to realize copyright authentication. According to the method, dynamic watermark embedding is achieved, the attack resistance is high, the image quality is well kept, an original model structure is not changed, and the method is suitable for copyright protection and user tracking of AIGC content.

This invention discloses a deep learning-based Hadamard transform-based anti-screen-capture watermarking method, belonging to the field of watermarking technology. This invention combines a convolutional neural network (CNN) and residual blocks to achieve an end-to-end process of watermark embedding and extraction within the Hadamard domain. By simulating screen-capture attacks and incorporating this method between the embedding and extraction layers, the network ensures robust watermark embedding. This method can efficiently extract watermark information from photos taken by surreptitiously, protecting the copyright of digital media. The use of the transform domain for watermark embedding allows the watermark to spread over a wider area of ​​the image, significantly improving the robustness of the watermarking algorithm. Comparison with related technologies demonstrates the superiority of this invention in terms of imperceptibility and robustness.

The application relates to the technical field of digital video processing, and discloses a video stream block time evidence watermarking method based on time authentication, which comprises the following steps: receiving a continuous video code stream and dividing the video code stream into multiple continuous video blocks, extracting video coding layer data of the first video block, performing a hash operation to generate a reference hash value; sending the reference hash value to a time watermarking device, performing a digital signature operation on the reference hash value in combination with an absolute timestamp and a signature key to generate a digital signature; intercepting a bit segment from the digital signature as a characteristic sequence to form cascade digest data; and modulating preset macro block syntax elements of the first frame of the next video block by using the cascade digest data, and outputting reconstructed video coding layer data. The method effectively reduces processor occupation and memory consumption of the system, enables video content to meet real-time transmission requirements in a monitoring scene while guaranteeing time sequence and data anti-fake capability.

The present invention proposes a hybrid video dual-watermarking method and system. The method extracts the chrominance channel of each video frame, calculates the extreme harmonic Fourier moment of the chrominance channel frame by frame, combines the calculation results with copyright information, and obtains a robust watermarked video frame. The robust watermarked video is then divided into various color channels. An enhanced least significant bit method is used for each color channel, combined with anti-counterfeiting information, to obtain a hybrid dual-watermarked video. The present invention can effectively protect valuable video content, accurately detect forged areas in the video, and demonstrate strong resistance to various attacks such as rotation, scaling, noise, and compression. In addition, the present invention can extract watermarks without the need for auxiliary information. The extreme harmonic Fourier moment of the chrominance channel of the transmitted video is calculated frame by frame to extract the robust watermark. The fragile watermark is extracted from each color channel of the video frame through an extraction function, and subsequent copyright authentication and anti-counterfeiting authentication operations can be completed.

This invention discloses a lightweight watermarking method for robust image attribution based on latent space diffusion models. The method uses low-dimensional feature vectors as the watermark carrier: first, a feature watermark library is constructed; then, a feature-noise encoder and a noise-feature decoder are trained. The encoder maps the feature watermark to initial latent space noise that satisfies a standard Gaussian distribution and inputs it into the latent space diffusion model to generate an image. When attribution verification is required, the image to be verified is inverted to recover the initial latent space noise, and the corresponding feature watermark is reconstructed by the decoder. The identity is determined using the cosine similarity between the reconstructed watermark and the original watermark in the feature watermark library. This invention maintains the quality of the generated image while exhibiting strong robustness to common distortions such as JPEG compression, cropping, random discarding, blurring, and brightness variations, and significantly reduces the overhead of watermark storage and matching computation.

The invention discloses a three-dimensional grid robust watermarking method and system based on spherical coordinate statistical characteristics, and the method comprises a watermark embedding stage and a watermark extraction stage, in the watermark embedding stage, three-dimensional grid vertexes are converted to a spherical coordinate space, a stable equivalent binning structure is constructed, and statistical characteristics are extracted, so that the three-dimensional grid robust watermarking is realized. And frequency domain transformation is carried out. In combination with a quantization domain binary mapping mode of quantization index modulation, watermark information is embedded into a frequency domain coefficient, and then a three-dimensional model carrying a watermark is generated through inverse transformation and coordinate reduction. The system comprises a coordinate conversion module, a binning module, a feature extraction module, a frequency domain transformation module, a watermark embedding module, an inverse transformation module, a vertex updating module and a model output module. The method provided by the invention has a large capacity up to 256 bits, and has significant robustness under rotation, scaling, translation, smoothing, simplification, vertex reordering and noise attack. The method can be widely applied to the technical field of three-dimensional data processing.

The invention relates to an official content credibility verification-oriented conditional trigger type text watermarking method, which comprises the following steps of: extracting confidence of a large language model, embedding a hierarchical trigger watermark, dividing a pseudo-random red and green set, detecting statistical characteristics and judging thresholding authenticity. Implicit watermark embedding of an official text and rapid discrimination of a public opinion end are realized. According to the method, the natural readability of the official text and the detectability of the watermark are both considered, the detection Z-score can reach 7.49, the green set hit proportion reaches 69.21%, and the method is suitable for the scenes of government affair document release, news draft clarification, network public opinion management and the like, and has good expansibility and engineering practicability.

The application discloses a large model watermarking method based on dynamic entropy, belonging to the field of large model watermarking. It includes: modifying the output token sampling process; dynamically adjusting the size of the sliding window and the size of the candidate token pool according to the probability distribution entropy value of the large model output; through pseudo-random scoring and multi-round competition; for each token in the sequence to be analyzed, the same watermark key and the context before the token are used to recalculate the pseudo-random seed, and based on the seed, the corresponding pseudo-random score of the token at all levels is regenerated; the average pseudo-random score of all tokens in the text is calculated as the aggregate score, and the statistical threshold is determined based on the preset false positive rate. If the statistical threshold is less than the aggregate score of the unknown text, it is determined as a generated text with watermark, otherwise it is a text without watermark. The application realizes high-precision source identification and tracing of large model output content through the dynamic entropy watermarking method while ensuring the original quality of the generated text.

The application discloses a kind of based on thumbnail encryption image's authenticatable user end watermark method, belong to computer security technical field. Including: image pre-processing, image encryption, key generation and distribution, decryption and watermark embedding, tamper detection and positioning, tamper content recovery, extract watermark lock user identity.The present application is based on the holographic redundancy embedding of thumbnail and the fine authentication based on parity check, with very strong anti-shearing and self-healing ability;Using the disturbance of pair of pixel difference value, all image pixels can be embedded watermark, and the security blind area is eliminated;Encrypted image has clear identifiable thumbnail, and is highly similar to plaintext, which is convenient for cloud management, and the details privacy is effectively protected through high-frequency information confusion;Encryption and watermark embedding are fused in bottom algebra operation, and the calculation efficiency is high.