A computer system for the creation of subliminal image or watermarks where the apparent video is an obvious image and a subliminal or clandestine image or watermark is hidden within the data structure. The concept is envisioned for the cloaking of images, sound, video or like digitized non-coded information. The cloaked files may be intended for storage, transmission, or clandestine placement in a public file system. The watermark may be used for the subliminal marking of a source of a file, its creation, or for tagging copyrighted information.

In one embodiment, a method including dividing a reference mask into a plurality of reference mask divisions, determining a plurality of motion vectors respectively associated with a plurality of slice divisions, wherein the plurality of reference mask divisions respectively correspond to the plurality of slice divisions, modifying a blurring kernel in accordance with the plurality of motion vectors, yielding a plurality of modified blurring kernels that are respectively associated with the plurality of slice divisions, and performing at least one action to yield an altered reference mask, including for the plurality of reference mask divisions and the plurality of modified blurring kernels: convolving a reference mask division with a weighted function of at least a modified blurring kernel associated with a slice division, of the plurality of slice divisions, to which the reference mask division corresponds.

The first image data is converted into the second image data defined by a color space that depends on an output apparatus that outputs image data. Additional information is multiplexed on the converted second image data by using a multiplex parameter.

Systems and methods are described for obfuscating variants of content segments. Variants of content segments can be used to encode an identifying sequence in a transmission of content. The variants of the content segments can each include one or more marked frames and one or more unmarked frames. Variations can be introduced into the unmarked frames for each of the variants of the content segments.

One aspect of the present invention discloses a watermark embedding method. The method includes: inputting an original bitstream; determining a first frame for embedding a watermark in the input original bitstream; selecting one of frames after the first frame as a second frame; generating the first frame a bidirectional-coded frame (B frame) referring to the second frame; generating the second frame as a reference frame of the first frame; and embedding the watermark in the first frame generated as the B frame.

Multi-blend fingerprinting may be detected. First, a video sample may be received. Next, frames of the received video sample may be step iteratively through until a probability value corresponding to a current frame indicates a match. Deciding that the probability value indicates the match may comprise creating an augmented frame, determining the probability value corresponding to the created augmented frame, and determining that the probability value indicates the match. Then a fingerprint from the created augmented frame may be extracted.

A method for encoding a digital signature of a neural network, implemented by an electronic device, the neural network being stored within a data structure including blocks of parameters. The method includes, for a current block of parameters including at least M parameters representing real numbers: an operation of acquiring a signature binary image from the digital signature including a set of bits extracted from parameters of the current block of parameters, and an operation of combining the signature binary image with a reference binary image delivering a coded binary image.

System, methods, and non-transitory computer-readable medium are provided for watermarking a diffusion model. For example, a method for watermarking a diffusion model may include generating one or more training data elements. In some aspects, the one or more trainings data elements may include target images. Moreover, the target images may include pre-defined watermark information. Further, the method may include training the diffusion model to predict the target images using training data including the one or more training data elements.

Methods, devices, and computer-program products are provided for adding and decoding data to a digital video signal in a visually imperceptible manner. For example, an encoded video frame can be obtained, and one or more blocks of the encoded video frame can be decoded. Binary data can be added to a subset of pixels from a set of pixels of the one or more blocks. For instance, a pixel component can be modulated to add the binary data. The one or more blocks can be re-encoded using at least one coding mode. The re-encoded one or more blocks can be added to the encoded video frame.

One aspect of the present invention discloses a watermark embedding method. The method includes: inputting an original bitstream; determining a first frame for embedding a watermark in the input original bitstream; selecting one of frames after the first frame as a second frame; generating the first frame a bidirectional-coded frame (B frame) referring to the second frame; generating the second frame as a reference frame of the first frame; and embedding the watermark in the first frame generated as the B frame.