Pith. sign in

REVIEW 23 cited by

A Sanity Check for AI-generated Image Detection

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2406.19435 v3 pith:7EE5VYS3 submitted 2024-06-27 cs.CV

Shilin Yan , Ouxiang Li , Jiayin Cai , Yanbin Hao , Xiaolong Jiang , Yao Hu , Weidi Xie This is my paper · ORCID
classification cs.CV
keywords ai-generatedimageimageschameleonachievesaideartifactsbenchmarks
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

With the rapid development of generative models, discerning AI-generated content has evoked increasing attention from both industry and academia. In this paper, we conduct a sanity check on "whether the task of AI-generated image detection has been solved". To start with, we present Chameleon dataset, consisting AIgenerated images that are genuinely challenging for human perception. To quantify the generalization of existing methods, we evaluate 9 off-the-shelf AI-generated image detectors on Chameleon dataset. Upon analysis, almost all models classify AI-generated images as real ones. Later, we propose AIDE (AI-generated Image DEtector with Hybrid Features), which leverages multiple experts to simultaneously extract visual artifacts and noise patterns. Specifically, to capture the high-level semantics, we utilize CLIP to compute the visual embedding. This effectively enables the model to discern AI-generated images based on semantics or contextual information; Secondly, we select the highest frequency patches and the lowest frequency patches in the image, and compute the low-level patchwise features, aiming to detect AI-generated images by low-level artifacts, for example, noise pattern, anti-aliasing, etc. While evaluating on existing benchmarks, for example, AIGCDetectBenchmark and GenImage, AIDE achieves +3.5% and +4.6% improvements to state-of-the-art methods, and on our proposed challenging Chameleon benchmarks, it also achieves the promising results, despite this problem for detecting AI-generated images is far from being solved.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 23 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. LEGO: LoRA-Enabled Generator-Oriented Framework for Synthetic Image Detection

    cs.CV 2026-05 unverdicted novelty 7.0 of 10

    LEGO uses multiple generator-specific LoRA modules modulated by an MLP and fused with attention to detect synthetic images, achieving better performance than prior methods while using under 10% of the training data.

  2. Bridging the Gap Between Ideal and Real-world Evaluation: Benchmarking AI-Generated Image Detection in Challenging Scenarios

    cs.CV 2025-09 conditional novelty 7.0 of 10

    AI-generated image detectors lose substantial accuracy on images shared over social media or scanned/re-photographed, while humans improve quickly after seeing two examples.

  3. Test-Time Curriculum for Open-Set AIGC Detection

    cs.CV 2026-08 conditional novelty 6.0 of 10

    A curriculum-based test-time adaptation method, using balanced confident pseudo-labels and multi-scale refinement, improves AIGC detector accuracy on unseen generators by 11 to 29 points over its starting detector.

  4. DECODE: Tackling Representation and Decision Degradation in Continual AI-Generated Image Detection

    cs.CV 2026-07 conditional novelty 6.0 of 10

    Continual fake-image detectors forget old generators through both feature drift and decision-boundary drift; DECODE's two-stage fix (diverse LoRA subspaces + closed-form head realignment) reports 99.36% accuracy, 0.39...

  5. LaP-Forensics: Latent-Pixel Consistency Guided Multimodal Reasoning for Deepfake Detection

    cs.CV 2026-07 conditional novelty 6.0 of 10

    A dual-stream deepfake forensic model that adds DDIM reconstruction residuals to RGB features improves artifact localization and cross-generator detection in evaluations, with honest caveats about text faithfulness.

  6. AI-generated Images Challenge Visual Trust in High-risk Scenarios

    cs.CV 2026-07 conditional novelty 6.0 of 10

    On SafeIMG, a new safety-focused benchmark of 1,131 GPT Image 2 images, the best VLM detects 49.5% of generated images and the best specialized detector 33.1%, versus 81.7% for humans.

  7. GlobalForge: Towards Robust AI-Generated Image Detection

    cs.CV 2026-07 conditional novelty 6.0 of 10

    GlobalForge improves AI-generated image detection under real-world degradation by suppressing local shortcuts and enforcing long-range structural reasoning, outperforming prior state-of-the-art by 5.89% average balanc...

  8. Continuously Evolving Deepfake Detection: An Architecture and Public-Benchmark Evaluation of a Dynamic Detection System

    cs.CV 2026-07 conditional novelty 6.0 of 10

    A continuously refreshed, incentive-driven deepfake detector beats static detectors on in-the-wild benchmarks and improves on post-export AI-generated media.

  9. VendorBench-100: A Unified Cross-Paradigm Benchmark for Deepfake Image Detection

    cs.CV 2026-07 conditional novelty 6.0 of 10

    A 36-model cross-paradigm benchmark on a hard 100-image corpus shows commercial APIs lead on MCC, open-source detectors trail on average, and a subset of strong rankers are miscalibrated at their default threshold.

  10. OC-Distill: Ontology-aware Contrastive Learning with Cross-Modal Distillation for ICU Risk Prediction

    cs.LG 2026-04 unverdicted novelty 6.0 of 10

    Ontology-aware contrastive pretraining plus note-to-vitals distillation improves MIMIC ICU risk and length-of-stay prediction using only vital signs at inference.

  11. AgentFoX: LLM Agent-Guided Fusion with eXplainability for AI-Generated Image Detection

    cs.CV 2026-03 conditional novelty 6.0 of 10

    An LLM agent guided by Expert and Clustering Profiles fuses heterogeneous AIGI detectors, resolves conflicts, and outputs explainable forensic reports that beat single experts and standard ensembles on high-conflict a...

  12. A Comprehensive Dataset for Human vs. AI Generated Image Detection

    cs.CV 2026-01 conditional novelty 6.0 of 10

    MS COCOAI provides 96,000 caption-aligned real and synthetic images from five generators, with baseline scores of about 0.80 for real-vs-AI detection and 0.45 for generator attribution.

  13. Sliced Transport Plans

    math.OC 2025-08 unverdicted novelty 6.0 of 10

    A new sliced-Wasserstein-based discrepancy with provable semi-metric properties and a constrained Kantorovich relation, plus a generic-measure generalization of Expected Sliced plans.

  14. AIGI-Holmes: Towards Explainable and Generalizable AI-Generated Image Detection via Multimodal Large Language Models

    cs.CV 2025-07 conditional novelty 6.0 of 10

    AIGI-Holmes combines visual expert pretraining, SFT on explanation data, and direct preference optimization to deliver human-verifiable explanations and top detection accuracy on unseen AI generators.

  15. Breaking Latent Prior Bias in Detectors for Generalizable AIGC Image Detection

    cs.CV 2025-06 conditional novelty 6.0 of 10

    Optimizing the initial latent noise of a diffusion model creates on-manifold adversarial examples that, added to training, make AIGC detectors generalize much better to unseen generators.

  16. OpenS2V-Nexus: A Detailed Benchmark and Million-Scale Dataset for Subject-to-Video Generation

    cs.CV 2025-05 conditional novelty 6.0 of 10

    A benchmark and five-million-clip dataset for evaluating and training subject-to-video generation models, with three new metrics for subject consistency, naturalness, and text alignment.

  17. Adaptive Classifier-Free Guidance via Dynamic Low-Confidence Masking

    cs.CL 2025-05 conditional novelty 6.0 of 10

    Adaptive Classifier-Free Guidance (A-CFG) re-masks low-confidence tokens in the unconditional input at each generation step, improving reasoning and planning accuracy for masked diffusion language models.

  18. Progressive Scaling Visual Object Tracking

    cs.CV 2025-05 reject novelty 6.0 of 10

    A progressive scaling training strategy with small-teacher distillation and masked-input alignment improves tracking accuracy and powers a new 12-dataset benchmark.

  19. Multimodal Conditional Information Bottleneck for Generalizable AI-Generated Image Detection

    cs.CV 2025-05 conditional novelty 6.0 of 10

    InfoFD detects AI-generated images by aligning compressed CLIP image features to orthogonalized text centroids, leveraging a layer-11 cosine-similarity bias between text and fake or real images.

  20. Leveraging Failed Samples: A Few-Shot and Training-Free Framework for Generalized Deepfake Detection

    cs.CV 2025-08 conditional novelty 4.0 of 10

    A training-free nearest-neighbor detector built on CLIP intermediate features uses a small number of labeled examples from each new generator to classify deepfakes, reporting strong few-shot accuracy across three benchmarks.

  21. LAID: Lightweight AI-Generated Image Detection in Spatial and Spectral Domains

    cs.CV 2025-07 conditional novelty 4.0 of 10

    A benchmark showing lightweight off-the-shelf models can exceed 99% accuracy on GenImage for AI-image detection, with spectral inputs and decision fusion, at a fraction of larger models' compute.

  22. NTIRE 2025 challenge on Text to Image Generation Model Quality Assessment

    cs.CV 2025-05 conditional novelty 4.0 of 10

    The NTIRE 2025 challenge report compares 20 methods for fine-grained text-to-image quality assessment, introduces the EvalMuse-Structure dataset, and finds every participating team outperformed the baselines.

  23. Survey on AI-Generated Media Detection: From Non-MLLM to MLLM

    cs.CV 2025-02 unverdicted novelty 3.0 of 10

    A survey organizing AI-generated media detection into Non-MLLM and MLLM based methods, with task and benchmark taxonomies.

Pith tools