IP Library › Granted Patent US 12,450,782

Granted Patent B2

US 12,450,782 · App. 18/471,456 · Granted Oct 21, 2025

Latent space based steganographic image generation

Inventors: Shruti Agarwal (Berkeley, CA); John Philip Collomosse (Pyrford, GB)

Assignee: Adobe Inc.

G06T9/002 G06T1/00 G06T1/0085 G06T2201/0053

View Patent ↗

Loading inventors, assignments & file history…

Monitor This Case

Get email alerts when status or documents change.

Order Certified Copies

Most orders are placed with the USPTO same day — all within 24 business hours.

Order via The Patent Place →

Pre-filled with this patent's details

Quick Facts

Patent No.: US 12,450,782
App. No.: 18/471,456
Granted: Oct 21, 2025
Kind: B2

External Links

USPTO Patent Center↗ Full Patent PDF↗

Abstract

Techniques for latent space based steganographic image generation are described. A processing device, for instance, receives a digital image and a secret that includes a bit string. A pretrained encoder of an autoencoder generates an embedding of the digital image that includes latent code. A secret encoder is trained and utilized to generate an embedding of the secret to act as a latent offset to the latent code. The processing device leverages a pretrained decoder of the autoencoder to generate a steganographic image based on the embedding of the secret and the embedding of the digital image. The steganographic image includes the secret and is visually indiscernible from the digital image. Further, the processing device is configured to recover the secret from the steganographic image, such as by training and leveraging a secret decoder to extract the secret.

Claims (33)

1. A method comprising:

receiving, by a processing device, a digital image and a secret that includes a bit string representing one or more characters;

generating, by a pretrained encoder of a convolutional neural network of the processing device, an embedding of the digital image that includes latent code;

generating, by a secret encoder of the processing device, an embedding of the secret to act as a latent offset to the latent code; and

outputting, by the processing device, a steganographic image that is visually indiscernible from the digital image and includes the secret, the steganographic image generated by inputting the embedding of the digital image and the embedding of the secret to a pretrained decoder of the convolutional neural network.

2. The method as described in claim 1 , wherein the embedding of the secret is generated with a dimensionality that corresponds to a dimensionality of the embedding of the digital image.

3. The method as described in claim 1 , wherein the embedding of the secret is incorporated into the latent code before input to the pretrained decoder of the convolutional neural network.

4. The method as described in claim 1 , wherein the secret includes content provenance information associated with the digital image.

5. The method as described in claim 1 , further comprising training the secret encoder using training pairs that each include a training digital image and a randomly generated training secret, the training further including generating training steganographic images based on the training pairs.

6. The method as described in claim 5 , wherein the training the secret encoder includes calculating a quality loss and a recovery loss.

7. The method as described in claim 6 , wherein the quality loss measures a visual difference between a particular training image and a corresponding training steganographic image and is based on a pixel loss and a perceptual loss.

8. The method as described in claim 1 , further comprising extracting the secret from the steganographic image using a secret decoder.

9. The method as described in claim 8 , wherein the secret decoder is trained to withstand image perturbations applied to the steganographic image using a noise model.

10. A system comprising:

a memory component; and

a processing device coupled to the memory component, the processing device to perform operations comprising:

receiving a steganographic image that includes a secret that includes a bit string representing one or more characters, the steganographic image generated by incorporating an embedding of the secret within latent code used to render the steganographic image;

extracting the secret from the steganographic image using a secret decoder, the secret decoder trained to withstand image perturbations using a noise model; and

outputting the one or more characters included in the secret in a user interface of the processing device.

11. The system as described in claim 10 , the operations further comprising training the secret decoder using training steganographic images and ground truth secrets to generate predicted secrets, the training includes determining a bit recovery loss based on the predicted secrets and corresponding ground truth secrets.

12. The system as described in claim 11 , wherein the training includes using the noise model to apply one or more image perturbations to the training steganographic images, the one or more image perturbations including one or more of a differentiable perturbation, a non-differentiable perturbation, or a perturbation that is approximatable with a differentiable transform.

13. The system as described in claim 12 , wherein the one or more image perturbations simulate redistribution of the steganographic image in an online context.

14. The system as described in claim 12 , wherein the one or more image perturbations include non-differentiable noise, and applying the one or more image perturbations includes converting the non-differentiable noise to additive noise.

15. The system as described in claim 10 , wherein the secret indicates whether the steganographic image was generated using generative artificial intelligence.

16. The system as described in claim 10 , wherein the secret indicates an authorship of the steganographic image.

17. A non-transitory computer-readable medium storing executable instructions, which when executed by a processing device, cause the processing device to perform operations comprising:

receiving a secret that includes a bit string representing one or more characters and latent code generated by a latent diffusion model;

generating, by a secret encoder of the processing device, an embedding of the secret to act as a latent offset to the latent code;

combining the embedding of the secret with the latent code generated by the latent diffusion model; and

outputting a steganographic image generated by a decoder of a convolutional neural network based on the combination of the embedding of the secret and the latent code, the secret visually imperceptibly hidden within the steganographic image.

18. The non-transitory computer-readable medium as described in claim 17 , wherein the latent code generated by the latent diffusion model is based on a text prompt.

19. The non-transitory computer-readable medium as described in claim 17 , wherein the latent code generated by the latent diffusion model is based on a Gaussian distribution.

20. The non-transitory computer-readable medium as described in claim 17 , wherein the operations further include extracting the secret from the steganographic image using a secret decoder.

Assignments (1)

ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 21, 2023

From: AGARWAL, SHRUTI; COLLOMOSSE, JOHN PHILIP

To: ADOBE INC.

Reel/Frame 064981/0631 →

Continuity (1)

Related Publication 20250104288A1 · Mar 27, 2025

References Cited (60)

US 20220405875A1 · Krishnamoorthy · 2022 [cited by examiner]

Baba, Sami et al., “Watermarking scheme for copyright protection of digital images”, IJCSNS International Journal of Computer Science and Network Security, vol. 9 No. 4 [retrieved Aug. 11, 2023]. Retrieved from the Inte… [cited by applicant]

Bharati, Aparna et al., “Transformation-aware embeddings for image provenance”, IEEE Transactions on Information Forensics and Security, vol. 16 [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://doi.org/10… [cited by applicant]

Black, Alexander et al., “Vpn: Video provenance network for robust content attribution”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/2109.10038.pdf>.… [cited by applicant]

Bose, R.C. et al., “On a class of error correcting binary group codes”, Information and Control, vol. 3, No. 1 [retrieved Aug. 11, 2023]. Retrieved from the Internet <doi.org/10.1016/S0019-9958(60)90287-4>., 1960, 12 Pa… [cited by applicant]

Caron, Mathilde et al., “Emerging Properties in Self-Supervised Vision Transformers”, Cornell University arXiv, arXiv.org [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/2104.14294.pdf>., M… [cited by applicant]

Ching-Chun, Chang et al., “Neural Reversible Steganography with Long Short-Term Memory”, Security and Communication Networks; London vol. 2021 [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://media.proque… [cited by applicant]

Coalition for Content Provenance, , “C2PA Technical Specification”, Coalition for Content Provenance Online Database [retrieved Nov. 29, 2023]. Retrieved from the Internet <https://c2pa.org/specifications/specifications… [cited by applicant]

Devi, P. et al., “A fragile watermarking scheme for image authentication with tamper localization using integer wavelet transform”, Journal of Computer Science, vol. 5, No. 11 [retrieved Aug. 11, 2023]. Retrieved from t… [cited by applicant]

Duan, Xintao , “Coverless information hiding based on generative model”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/1802.03528.pdf>., Feb. 10, 2018,… [cited by applicant]

Duan, Xintao et al., “Reversible Image Steganography Scheme Based on a U-Net Structure”, IEEE Access, vol. 7 [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://web.archive.org/web/20210429085409id_/https://… [cited by applicant]

Esser, Patrick et al., “Taming Transformers for High-Resolution Image Synthesis”, Cornell University arXiv, arXiv.org [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/2012.09841.pdf>., Jun. … [cited by applicant]

Fernandez, Pierre et al., “Watermarking images in selfsupervised latent spaces”, Cornell University arXiv, arXiv.org [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/2112.09581.pdf>., Mar. 2… [cited by applicant]

Fernandez, Pierre et al., “Watermarking images in self-supervised latent spaces”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/2112.09581.pdf>., Mar. … [cited by applicant]

Ghazanfari, Kazem et al., “Lsb++: An improvement to Isb+ steganography”, Tencon IEEE Region 10 Conference [retrieved Aug. 10, 2023]. Retrieved from the Internet <10.1109/TENCON.2011.6129126>., Nov. 2011, 5 Pages. [cited by applicant]

He, Kaiming et al., “Deep Residual Learning for Image Recognition”, Cornell University arXiv Preprint, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/1512.03385.pdf>., Dec. 10, 2… [cited by applicant]

Hendrycks, Dan et al., “Benchmarking neural network robustness to common corruptions and perturbations”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/… [cited by applicant]

Hodosh, M. et al., “Framing Image Description as a Ranking Task: Data, Models and Evaluation Metrics”, Journal of Artificial Intelligence Research, vol. 47 [retrieved Aug. 11, 2023]. Retrieved from the Internet <https:/… [cited by applicant]

Holub, Vojtěch et al., “Designing steganographic distortion using directional filters”, IEEE International Workshop on Information Forensics and Security (WIFS) [retrieved Aug. 10, 2023]. Retrieved from the Internet <ht… [cited by applicant]

Holub, Vojtěch et al., “Universal distortion function for steganography in an arbitrary domain”, EURASIP Journal on Information Security [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://citeseerx.ist.psu.… [cited by applicant]

Huang, Chen-Hsiu , “Image data hiding with multi-scale autoencoder network”, Cornell University arXiv, arXiv.org [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/2201.06038.pdf>., Jan. 16, 2… [cited by applicant]

Huiskes, Mark J. et al., “The mir flickr retrieval evaluation”, MIR '08: Proceedings of the 1st ACM international conference on Multimedia information retrieval [retrieved Aug. 11, 2023]. Retrieved from the Internet <ht… [cited by applicant]

Karras, Tero et al., “Training Generative Adversarial Networks with Limited Data”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/2006.06676.pdf>., Oct.… [cited by applicant]

Li, Ming et al., “Improve Image Codec's Performance By Variating Post Enhancing Neural Network:Submission of zxw for CLIC2020”, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Wo… [cited by applicant]

Li, Xiaoxia et al., “A steganographic method based upon JPEG and particle swarm optimization algorithm”, Information Sciences: an International Journal, vol. 177, No. 15 [retrieved Aug. 10, 2023]. Retrieved from the Int… [cited by applicant]

Liu, Xiyao et al., “Image disentanglement autoencoder for steganography without embedding”, IEEE/CVF Conference on Computer Vision and Pattern Recognition [retrieved Aug. 11, 2023]. Retrieved from the Internet <10.1109/… [cited by applicant]

Meng, Ruohan et al., “A fusion steganographic algorithm based on faster ronn”, Computers, Materials & Continua vol. 55, No. 1 [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://cdn.techscience.cn/files/cmc/… [cited by applicant]

Meng, Ruohan , “A Fusion Steganographic Algorithm Based on Faster R-CNN”, Computers, Materials & Continua . 2018, vol. 55 No. 1 [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://cdn.techscience.cn/files/cm… [cited by applicant]

Navas, K. A. et al., “DWT-DCT-SVD Based Watermarking”, 3rd International Conference on Communication Systems Software and Middleware and Workshops [retrieved Nov. 29, 2023]. Retrieved from the Internet <https://doi.org/… [cited by applicant]

Nguyen, E. et al., “Oscar-net: Object-centric scene graph attention for image attribution”, IEEE International Conference on Computer Vision [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://openaccess.the… [cited by applicant]

Pevný, Tomáš et al., “Using High-Dimensional Image Models to Perform Highly Undetectable Steganography”, IH'10: Proceedings of the 12th international conference on Information hiding [retrieved Aug. 19, 2023]. Retrieved… [cited by applicant]

Provos, Niels , “Defending against statistical steganalysis”, Electrical Engineering and Computer Science, Department of (EECS) Engineering [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://deepblue.lib.um… [cited by applicant]

Qin, Jiaohua et al., “Coverless image steganography: a survey”, IEEE Access, vol. 7 [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8911367>., Nov. 25, 20… [cited by applicant]

Ramachandran, Prajit et al., “Swish: a self-gated activation function”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/1710.05941v1.pdf?source=post_page… [cited by applicant]

Ramesh, Aditya et al., “Hierarchical Text-Conditional Image Generation with CLIP Latents”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/2204.06125.pdf… [cited by applicant]

Rombach, Robin et al., “High-Resolution Image Synthesis with Latent Diffusion Models”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/2112.10752.pdf>., … [cited by applicant]

Saharia, Chitwan et al., “Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pd… [cited by applicant]

Selvaraju, Ramprasaath et al., “Grad-cam: Visual explanations from deep networks via gradient-based localization”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxi… [cited by applicant]

Shaham, Tamar R. et al., “SinGAN: Learning a Generative Model From a Single Natural Image”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2022]. Retrieved from the Internet <https://arxiv.org/pdf/1905.01164.pd… [cited by applicant]

Shin, Richard et al., “JPEG-resistant Adversarial Images”, NIPS 2017 Workshop on Machine Learning and Computer Security, vol. 1 [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://machine-learning-and-securi… [cited by applicant]

Subramanian, Nandhini et al., “End-to-End Image Steganography Using Deep Convolutional Autoencoders”, IEEE Access, vol. 9 [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://irep.ntu.ac.uk/id/eprint/46088/1/… [cited by applicant]

Taha, Mustafa Sabah et al., “High payload image steganography scheme with minimum distortion based on distinction grade value method”, Multimedia Tools and Applications, vol. 81, No. 18 [retrieved Aug. 10, 2023]. Retrie… [cited by applicant]

Tancik, Matthew , “StegaStamp: Invisible Hyperlinks in Physical Photographs”, Cornell University arXiv, arXiv.org [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/1904.05343.pdf>., Mar. 26, … [cited by applicant]

Volkhonskiy, Denis et al., “Steganographic generative adversarial networks”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/1703.05502.pdf>., Oct. 7, 20… [cited by applicant]

Wan, Wenbo et al., “A comprehensive survey on robust image watermarking”, Neurocomputing, vol. 488, No. C [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://pure.port.ac.uk/ws/files/50956558/Survey_on_Water… [cited by applicant]

Weng, Xinyu et al., “High-Capacity Convolutional Video Steganography with Temporal Residual Modeling”, ICMR '19: Proceedings of the 2019 on International Conference on Multimedia Retrieval [retrieved Nov. 7, 2023]. Retr… [cited by applicant]

Westfeld, Andreas et al., “F5—A Steganographic Algorithm”, In: Moskowitz, I.S. (eds) Information Hiding. IH 2001. Lecture Notes in Computer Science, vol. 2137. Springer, Berlin, Heidelberg. <https://doi.org/10.1007/3-54… [cited by applicant]

Wolfgang, R.B. et al., “A watermark for digital images”, Proceedings of 3rd IEEE International Conference on Image Processing [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://www.cerias.purdue.edu/assets/… [cited by applicant]

Wu, Pin et al., “StegNet: Mega Image Steganography Capacity with Deep Convolutional Network”, Cornell University arXiv, arXiv.org [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/1806.06357.… [cited by applicant]

Yu, Ning et al., “Artificial Fingerprinting for Generative Models: Rooting Deepfake Attribution in Training Data”, Cornell University, arXiv Preprint, arxiv.org [retrieved Nov. 7, 2023]. Retrieved from the Internet <htt… [cited by applicant]

Yu, Ning et al., “Responsible Disclosure of Generative Models Using Scalable Fingerprinting”, Cornell University, arXiv Preprint, arxiv.org [retrieved Nov. 10, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/2… [cited by applicant]

Zhang, Kevin Alex et al., “Robust Invisible Video Watermarking with Attention”, Cornell University arXiv, arXiv.org [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/1909.01285.pdf>., Sep. 3,… [cited by applicant]

Zhang, Lvmin et al., “Adding Conditional Control to Text-to-Image Diffusion Models”, Cornell University, arXiv Preprint, arxiv.org [retrieved Nov. 10, 2023]. Retrieved from the internet <https://arxiv.org/pdf/2302.05543… [cited by applicant]

Zhang, Richard et al., “The Unreasonable Effectiveness of Deep Features as a Perceptual Metric”, Cornell University arXiv, arXiv.org [retrieved Aug. 11, 2023]. Retrieved from the Internet <https://arxiv.org/pdf/1801.039… [cited by applicant]

Zhang, Xu et al., “Discovering Image Manipulation History by Pairwise Relation and Forensics Tools”, IEEE Journal of Selected Topics in Signal Processing, vol. 14, No. 5 [retrieved Nov. 28, 2023]. Retrieved from the Int… [cited by applicant]

Zhou, Zhi Li et al., “Coverless Information Hiding Based on Bag-of-Words Model of Image”, Journal of Applied Sciences, vol. 34, No. 5 [retrieved Nov. 28, 2023]. Retrieved from the Internet <https://doi.org/10.3969/j.iss… [cited by applicant]

Zhou, Zhili et al., “Coverless Image Steganography Without Embedding”, In: International Conference on Cloud Computing and Security [retrieved Nov. 28, 2023]. Retrieved from the Internet <https://doi.org/10.1007/978-3-3… [cited by applicant]

Zhou, Zhili et al., “Encoding multiple contextual clues for partial-duplicate image retrieval”, Pattern Recognition Letters, vol. 109 [retrieved Dec. 14, 2023]. Retrieved from the Internet <https://doi.org/10.1016/j.pat… [cited by applicant]

Zhu, Jiren et al., “HiDDeN: Hiding Data with Deep Networks”, Proceedings of the European Conference on Computer Vision [retrieved Aug. 10, 2023]. Retrieved from the Internet <https://openaccess.thecvf.com/content_ECCV_2… [cited by applicant]

Cited By (1)

US 12,712,706