Integrating Wavelet Feature Decomposition and 3D CNNs for Accurate Blind Steganalysis
Bibliographic record
Abstract
Blind steganalysis aims to determine whether a piece of media possesses hidden information without prior knowledge of the embedding algorithm.This task has become increasingly challenging as steganographic techniques continue to evolve rapidly.In this paper, we present a novel approach that integrates wavelet-based feature representations with a three-dimensional deep convolutional neural network (3D CNN) for robust blind image steganography.The discrete wavelet transform (DWT) is employed to capture spatial-frequency characteristics across subbands, enabling the preservation of subtle embedding distortions that conventional spatial-domain approaches often overlook.These wavelet-based feature volumes serve as inputs to the 3D CNN, which jointly models interband, spatial, and frequency-domain dependencies through volumetric convolution.To rectify class imbalance and increase classification robustness, we introduce a custom weighted classification layer.We conducted extensive experiments on the BOWS2 and BOSSBase v1.01 datasets, and the results demonstrate that the proposed method outperforms baseline models using 2D CNN architectures in terms of accuracy, precision, recall, and F1-score across all embedding schemes.Our results demonstrate the potential of combining wavelet-domain methods with volumetric deep learning (DL) to improve blind steganalysis in practical digital forensics and cybersecurity applications.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".