A data-generation pipeline that captions real photographs with a vision–language model and regenerates them with modern text-to-image systems, producing semantically aligned real/synthetic pairs that isolate generative artifacts from image content is described.
Abstract
We study the detection of AI-generated images and contribute toward detectors that are accurate, trustworthy, and honestly evaluated. We first describe a data-generation pipeline that captions real photographs with a vision–language model and regenerates them with modern text-to-image systems, producing semantically aligned real/synthetic pairs that isolate generative artifacts from image content. We then build a lightweight, CPU-deployable spectral detector that fuses an RGB backbone with a radial residual-frequency branch, and show through a controlled ablation that the frequency cue mainly contributes calibration and false-positive control rather than raw ranking. To keep a single saturated score from overstating readiness, we package the detector with a multi-axis evaluation suite and a worst-group summary metric. Finally, we extend the same recipe to a dedicated face-deepfake detector that adds a neighboring-pixel-relationship branch and generalizes well to generators unseen in training. We release the models, the data protocols, and all per-sample scores as a reproducible reference point for synthetic-image forensics.
GurAI is proposed, a transparent logistic late-fusion method that combines Rich384 and DeMamba logits that suggests that transparent late fusion can exploit complementary detector strengths more effectively than architectural redesign alone when facing generator diversity.
GenSyn10 is introduced, a CIFAR-10-aligned synthetic image dataset of 60,000 images generated using three architecturally diverse state-of-the-art models, enabling controlled and systematic evaluation of out-of-distribution (OOD) generalization to novel generators.
Md Faraz Kabir Khan, Saeed Anwar, G. Hassan· 0 citations
Across cross-generator, post-processing, and in-the-wild benchmarks, PE-SPC surpasses the previous DINOv3 baseline and achieves new state-of-the-art results.
Weihan Cai, Hao Tan, Zichang Tan et al.· 0 citations
Deepfake image detection is served by three fundamentally different paradigms - commercial APIs, zero-shot vision-language models (LLMs), and open-source detectors - that are rarely evaluated under a common protocol, making direct comparison difficult. We introduce VendorBench-100, a cross-paradigm benchmark that evaluates 36 representative models using a single adversarial 100-image corpus, a unified output schema, and a common evaluation framework. Models are ranked primarily by the Matthews correlation coefficient (MCC), with ROC-AUC reported as a threshold-independent measure of ranking ability. Rather than maximizing size, it emphasizes real-world difficulty through a taxonomy of eight edge-case families such as face swaps, text-to-video stills, AI photo edits, avatar compositing, opaque-provenance images, and compressed research frames. Commercial APIs achieve the strongest median performance, followed by vision LLMs and open-source detectors, though individual open-source models remain competitive with the best LLMs. Across all 36 models, MCC and ROC-AUC are strongly correlated (Pearson r ~ 0.86); the more consequential finding is narrower and one-directional: a subset of otherwise strong rankers are miscalibrated at their shipped default threshold, so a high ROC-AUC can overstate real-world deployability. Separately, raw accuracy and F1 are unreliable on this corpus's imbalanced class split, since a model that predicts"fake"indiscriminately scores deceptively well on both while offering no real discriminative skill. No single metric is safe in isolation: MCC and specificity should always accompany ROC-AUC and accuracy. We release the complete evaluation framework and results. Code and data: https://github.com/sharayu-20/vendorbench-100
S. N. Deshmukh, Md. Rashidunnabi, Nelton Tiago Gemo et al.· 0 citations
The rapid advancement of text-to-image (T2I) models has necessitated robust Synthetic Image Source Attribution (SIA) methodologies. A critical challenge in SIA is the distribution shift between pristine training images and real-world deployed images, which undergo unknown post-processing operations such as JPEG compression and blurring. In this work, proposed for the DLMMDD Challenge at ICANN 2026, we introduce a dual-branch ensemble framework fusing Semantic Deep Learning with Mathematical Forensic Feature Extraction. The semantic branch employs EfficientNet-B0 regularized with Exponential Moving Averaging (EMA) and Label Smoothing. The forensic branch extracts 126 mathematical features -- including SVD spectral profiles and Local Binary Patterns -- from high-pass noise residuals, compressed via Truncated SVD and classified with XGBoost. Evaluated on a dataset of 10 generators where 55% of the test set is degraded, our approach achieves a private leaderboard accuracy of 95.60%. Furthermore, the entire pipeline is highly computationally efficient, requiring no GPU acceleration and executing end-to-end on a standard CPU in under 6.5 hours, highlighting the practicality and scalability of mathematical forensics for real-world deployment.
The rapid advancement of generative AI has enabled the creation of highly realistic deepfake media, posing significant threats, including misinformation, digital identity theft, fraud, and manipulation of public opinion. AI-generated image (AIGI) detection is reliably challenging due to the diversity of generative methods and the subtle artifacts they leave behind. In this work, we propose GenRes, a novel framework for generative residual learning via a neural tensor network, which models fine-grained relational features between original and transformed samples to enhance generalization. To address scenarios involving multiple generative transformations, we introduce GenRes++, which employs a learnable attention mechanism to aggregate relational features across multiple transformed samples and enables the model to focus on the most informative cues. Both models leverage PE-Core as a feature extractor, providing generalized and semantically rich embeddings that improve cross-domain performance and enable the detection of AIGI generated by unseen methods. Comprehensive experiments on multiple benchmark datasets demonstrate that the proposed GenRes++ approach outperforms existing methods.
Kutub Uddin, Nusrat Tasnim, Awais Khan et al.· 2 citations