Fusing Frequency and Spatial Transformers for Robust Detection of AI-Generated Images
Bibliographic record
Abstract
The rapid advancement of generative models, particularly Generative Adversarial Networks (GANs), has led to the proliferation of highly realistic AI-generated images that pose serious challenges to digital content authenticity.This paper presents a dual-branch transformerbased architecture designed to enhance the detection of such synthetic images by simultaneously learning spatial and frequency-domain representations.The proposed model processes RGB inputs and their corresponding Fast Fourier Transform (FFT)-based spectrograms through two parallel Vision Transformer encoders, enabling the extraction of complementary features.These features are fused before final classification, allowing the model to capture both local texture inconsistencies and global signal anomalies that are characteristic of AI-generated imagery.The system was evaluated on a dataset comprising real and StyleGAN2-generated facial images, trained on real and StyleGAN2-generated face images, the model achieved a validation AUC of 0.9807 and generalized effectively to unseen StyleGAN3 samples.An ablation study confirmed the contribution of the frequency stream, and additional testing on StyleGAN3-generated images-unseen during trainingdemonstrated the model's strong generalization capability.These findings suggest that combining spectral and spatial learning within a Transformer framework offers a robust solution for detecting AI-synthesized images in increasingly complex visual environments.
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 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| 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".