Robust and Secure Image Fingerprinting Learned by Neural Network
Abstract
Image fingerprinting is a technique that summarizes the perceptual characteristics of a digital image into an invariant digest, and it is one of the most effective solutions for digital rights management. Most conventional fingerprinting algorithms were developed by assembling manually designed feature extractor and quantizer, which requires extensive expert knowledge and may not capture the intrinsic or abstract visual characteristics of the digital image. Focusing on content identification related applications, we propose a data-driven image fingerprinting algorithm in this paper, where neural network is trained to automatically discover the optimal mapping from image to fingerprint. To ameliorate the difficulty of training, we start by training the fingerprint-computation network in a layer-wise manner to progressively improve its robustness against content-preserving distortions. Initialized by the states learned by layer-wise training, the network is then re-trained as a holistic unit, with the objective of maximizing its content identification accuracy. Moreover, we also develop a key-dependent version of the neural network-based fingerprinting algorithm. By quantifying its security using information-theoretic metrics, we have proved that the hierarchical architecture of neural network is beneficial to the security of fingerprinting algorithm. The experimental results on a large testing database show that the proposed work exhibits much higher content identification accuracy than state-of-the-art algorithms, and its execution speed is in the millisecond time scale.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Some deep learning schemes consider a limited variety of content retention operations, which is not enough to deal with image copyright protection.
“Some schemes based on deep learning are designed to consider a limited variety of content retention operations, which are not enough to deal with the increasingly severe situation of image copyright protection.”
From Deep Perceptual Hash Based on Hash Center for Image Copyright Protection (2022) · cites this paper as [33]
CNMC and IFNN are specially designed for picture copyright protection.
“Where CNMC and IFNN are specially designed for picture copyright protection.”
From Deep Perceptual Hash Based on Hash Center for Image Copyright Protection (2022) · cites this paper as [33]
CNMC, IFNN, DSH, and DBDH are based on deep learning.
“CNMC [34], IFNN [33], DSH [29] and DBDH [44] are based on deep learning.”
From Deep Perceptual Hash Based on Hash Center for Image Copyright Protection (2022) · cites this paper as [33]
There is few research on generating perceptual image hashes using deep neural networks, and there are many shortcomings, such as the large number of original images required for training and the insufficient variety of transformations that can be resisted.
“However, to the best of our knowledge, there is few research on generating perceptual image hashes using deep neural networks, and there are many shortcomings, such as the large number of original images required for training and the insuf cient variety of transformations that can be resisted.”
From Deep Perceptual Hash Based on Hash Center for Image Copyright Protection (2022) · cites this paper as [33]
Li et al.'s method [33] takes the lead in applying deep learning to the field of Perceptual Hashing, but it is not an end-to-end method and the combination of multiple codecs are still manually designed.
“This method [33] takes the lead in applying deep learning to the eld of Perceptual Hashing, but it is not an end-to-end method and the combination of multiple codecs are still manually designed.”
From Deep Perceptual Hash Based on Hash Center for Image Copyright Protection (2022) · cites this paper as [33]
The curves of DSH, CNMC, and IFNN schemes are very close.
“The curves of DSH [29], CNMC [34] and IFNN [33] schemes are very close.”
From Deep Perceptual Hash Based on Hash Center for Image Copyright Protection (2022) · cites this paper as [33]
Cites
Perceptual Image Hashing Via Feature Points: Performance Evaluation and Tradeoffs
Monga, V., & Evans, B. L. (2006) · IEEE Transactions on Image Processing · doi:10.1109/tip.2006.881948 · Cited by 363 · cited 1× · see citing works
“Some local and motion descriptors, like key point [15]–[17], gradient orientations [18] and optical flow [19], have also been explored for designing image and video fingerprinting algorithms.”
Background
Flagged first, then by citations
Cited by
Sorted by citations
Qin, C., Liu, E., Feng, G., & Zhang, X. (2021) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2020.3047142 · Cited by 78 · see citing works
Unified Performance Evaluation Method for Perceptual Image Hashing
Li, X., Qin, C., Wang, Z., Qian, Z., & Zhang, X. (2022) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2022.3161149 · Cited by 32 · see citing works
Deep Perceptual Hash Based on Hash Center for Image Copyright Protection
Sun, X., & Zhou, J. (2022) · IEEE Access · doi:10.1109/access.2022.3221980 · Cited by 20 · see citing works
“This method [33] takes the lead in applying deep learning to the field of Perceptual Hashing, but it is not an end-to-end method and the combination of multiple codecs are still manually designed.”
BackgroundMethodInfluentialPerceptual authentication hashing for digital images based on multi-domain feature fusion
Cao, F., Yao, S., Zhou, Y., Yao, H., & Qin, C. (2024) · Signal Processing · doi:10.1016/j.sigpro.2024.109576 · Cited by 8 · see citing works
A High-Performance Region Recognition Network-Enhanced Deep CNN for Image Content Perceptual Hashing
Yang, M., Qi, B., Ma, B., Xu, J., Xian, Y., & Li, X. (2025) · IEEE Internet of Things Journal · doi:10.1109/jiot.2025.3559675 · Cited by 6 · see citing works
Document-Image Perceptual Hashing for Content Authentication
Situ, X., Liu, T., Yao, H., Qin, C., & Zhang, X. (2025) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2025.3596854 · Cited by 2 · see citing works
Perceptual Robust Hashing for Video Copy Detection with Unsupervised Learning
Zhao, G., Qin, C., Luo, X., Zhang, X., & Chang, C.-C. (2023) · Proceedings of the 2023 ACM Workshop on Information Hiding and Multimedia Security · doi:10.1145/3577163.3595110 · Cited by 1 · see citing works
- Authors
- Li, Y., Wang, D., & Tang, L.
- Affiliations
- Tianjin University (CN)
- Harbin Institute of Technology (CN)
- Venue
- IEEE Transactions on Circuits and Systems for Video Technology
- Published
- 2020-02
- Volume
- 30
- Issue
- 2
- Pages
- 362-375
- ISSN
- 1051-8215 (ISSN-L)
- 1558-2205
- OpenAlex
- W2911013958
- Cited by
- 52
- References
- 1
- Citations by year
- 2026: 4 · 2025: 9 · 2024: 10 · 2023: 11 · 2022: 11 · 2021: 4 · 2020: 3
- FWCI
- 2.02
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- IEEE Transactions on Circuits and Systems for Video Technology (publishedVersion)
- Funders
- National Natural Science Foundation of China (61572352, 61202164)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Steganography and Watermarking Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Retrieval and Classification Techniques
- Keywords
- image fingerprinting, digital rights management, content identification, layer-wise training, data-driven feature extraction
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
