CS Paper Radar

SpectralCanary: Embarrassingly Simple Low-Frequency Traces for T2I Dataset-Use Auditing

Sy-Tuyen Ho, Anirudh Satheesh, Soham Chawathe, Huy Nghiem, Yongkyun Lee, Andrew Adams Schoen, Furong Huang

cs.CV cs.LG ·submitted 2026-10-22 17:59 UTC · v1
published

Abstract

Auditing whether a text-to-image (T2I) model was fine-tuned on a particular image collection is difficult: a model can absorb collection-level visual characteristics without reproducing training images. Proactive dataset-use auditing plants a shared signal in a dataset before release, then later checks whether that signal appears in a suspect model's generated outputs. Existing approaches often trade off verifiability, fidelity, robustness, and deployability. We introduce $\textbf{SpectralCanary}$, which uses a smooth low-pass mask to blend a fixed canary image into every collection image while leaving captions unchanged and requiring neither auxiliary generation nor per-image optimization. A lightweight verifier detects the transferred signal in suspect-model outputs. Across eight tests on the Pokemon dataset and three realistic data transformations, spanning two architectures families, SpectralCanary achieves the highest average per-generation TPR@$1\%$ FPR ($91.1\%$) and fixed threshold model-level accuracy using 30 generations ($94.8\%$) among the evaluated methods. Across five datasets spanning artistic, medical, satellite, and fashion domains with SD3.5, SpectralCanary averages $92.4\%$ per-generation TPR@$1\%$ FPR while preserving the released images (SSIM $0.989$, LPIPS $0.022$) with a near-zero mean change FID (mean $Δ$FID $0.05$) across datasets. Matched-energy ablations on Pokemon and ROCOv2 support our design choice, low-frequency placement transfers more reliably compared to mid/high-frequency placement. Under the evaluated adaptive attacks, the most effective removal requires aggressive modification and incurs more than $25$ $Δ$FID points. These results identify low-frequency visual signals as effective carriers for collection-level T2I audit traces.

Comment
NeurIPS 2026 Workshop on Attributing Model Behavior at Scale

Closest papers

Nearest neighbours by meaning across the whole index.

Verifier Choice is a Benchmark Design Variable: Auditing Structural Counting Evaluation in Text-to-Image Models

cs.CV cs.LG ·2026-11-11 ·Shurun Li, Michael Hua Wang, Haibo Zeng, Long Wang
top venuebenchmarkpositioncode/datahot direction similarity 0.63
Abstract · Accepted at NeurIPS 2026 (Evaluations and Datasets Track). 24 pages. Data: https://huggingface.co/datasets/neurips2026-ed443/audit-release .…

Automatic evaluation is now central to text-to-image benchmarking, but the verifier that converts generated images into scores is often treated as an implementation detail. We argue that this is unsafe: an automatic benchmark score is a property of the image generator, prompt distribution, verifier, verifier-use protocol, and scoring rule, not of the generator alone. We study this issue in structural counting, a controlled setting where human-visible counts can be audited. We construct a verifier-centric audit suite over seven counting objects: cell phones, chairs, and traffic lights for object counting, and clover leaflets, dresser drawer fronts, power-strip sockets, and turbine blades for substructure counting. Across 6,240 generated images from Qwen-Image-2512 and FLUX.2-dev, we compare open MLLM verifiers under count-only, reject-aware, zero-shot and few-shot protocols, plus TIFA-like target queries, a detector baseline, and frontier closed-model checks on two hard objects. On a balanced 2,280-image evaluation split, aggregate score matching can coexist with image-level disagreement, and verifier-use protocols and generator distributions change the measured reliability. Every verifier-protocol configuration also shrinks the human score difference between the two generators, by 7.5-25.0 pp, and with a third generator, Z-Image, all four tested protocols reverse the human ordering of Z-Image and FLUX.2-dev on our fixed prompt suite. GPT-5.5 and Claude Opus 4.7 checks do not remove the failure modes: GPT-5.5 still has only 59.92% clover valid-count fidelity, while Claude Opus 4.7 has lower same-slice agreement ($κ=0.432$). We propose reporting automatic benchmark scores with a factorized verifier reliability profile: aggregate score gap, image-level agreement, generator-comparison distortion, valid-image count fidelity, rejection behavior, output compliance, and directional bias.

0.57

MemCatalyst: Amplifying Data Auditing on Vision-Language Models via Data Poisoning

cs.CR cs.LG ·2026-08-18 ·Xukun Luan, Jinyan Liu, Yuhui Gong, Yuanguo Bi, Bing Hu +2 more
similarity 0.60
Abstract

Vision-Language models (VLMs) achieve outstanding performance largely due to the amount of training data available on the internet. At the same time, data holders (e.g., artists) urgently need to determine whether their data has been used for model training without authorization, which concerns both intellectual property rights and personal privacy. Data auditing, particularly through membership inference (MI), has attracted attention as a direct tool. This work proposes MemCatalyst, a set of data poisoning tools, aiming to amplify the data auditing performance on VLMs. MemCatalyst employs two strategies: Poisoning Text (PT) and Poisoning Image (PI). MemCatalyst forces VLMs to over-learn specific inconsistencies between image features and textual semantics during training, thereby increasing their susceptibility to membership information auditing. Crucially, the transferability of poisoned samples across different VLM architectures is demonstrated to be effective in the black-box setting. Extensive evaluations using five state-of-the-art data audits on two prominent VLMs demonstrate that MemCatalyst markedly enhances MI AUC scores with a minimal budget of poisoned samples, while maintaining a negligible impact on model performance.

0.17

CropCop: An Auditable 120-Class Plant-Health Model from Benchmark Reconstruction to a Quantised Runtime Artifact

cs.CV cs.LG ·2026-08-26 ·Rana Muhammad Ahmed, Sabahat Abbas
benchmark similarity 0.60
Abstract · 25 pages, 6 figures, and 15 tables. Includes benchmark-audit, model-retention, runtime-fidelity, and reproducibility appendices

A plant-health score can appear precise while resting on duplicated image families, a long-tailed label space, or a runtime file that was never evaluated. We present CropCop, a closed-set recognition system spanning 120 operational plant-health classes and an evidence chain from corpus reconstruction to direct execution of the final quantised artifact. Starting from 117,546 audited images, we rejected the inherited partition after confirming 3,233 duplicate relationships across split boundaries and froze a 109,107-image benchmark with zero crossings among the audited trusted leakage groups and a 151.7 largest-to-smallest class ratio. A fully fine-tuned DINOv3 ConvNeXt-Tiny reference achieved 98.51% accuracy and 96.87% macro-F1 on the locked internal test. A compact MobileNetV4 Conv-Medium derivative achieved 98.46% accuracy and 96.27% macro-F1 without being presented as evidence for a new distillation method. Validation-only post-training quantisation selected dynamic activations with per-channel weights, and the final 22.60 MiB ExecuTorch/XNNPACK PTE achieved 98.46% accuracy and 96.23% macro-F1 when executed directly. Only six of 16,363 top-1 decisions changed between the converted INT8 graph and the PTE, while paired analysis showed a modest class-balanced loss; an exploratory post hoc fruit-label slice localized a larger recall decline than aggregate accuracy revealed. CropCop establishes strong leakage-controlled internal recognition and software-runtime fidelity; it does not establish performance on unseen farms, camera pipelines, or physical Android hardware.

0.38

VTC-AUDIT: An Interventional Reliability Audit for Visual-Token Compression under Paired Image Shifts

cs.CV ·2026-10-25 ·Isaac, Li
benchmarkcode/data similarity 0.59
Abstract · 9 pages, 6 tables. Code and experimental results: https://github.com/IsaacLi74/VTC-AUDIT

Visual-token compression is usually evaluated by clean accuracy and efficiency, but these averages do not reveal failures that compression introduces under an image shift where the uncompressed model stays correct. We introduce VTC-AUDIT, which estimates this incremental risk only on pairs where the full model is correct on both images and the compressed model is correct on the clean image. The four arms then cross two binary choices: which image selects the retained token positions, and which image supplies the visual features. We audit three open 8-9B vision-language models on four tasks and eight pre-specified shifts. We evaluate two training-free compressors---hard-pruning DivPrune and prune-and-merge ApET---at a matched 25% retained-position budget, producing 708,000 inference records. Reselecting positions on the shifted image, rather than reusing the clean-image positions, raises the failure rate in all six model-compressor settings. The increase is 3.04-9.60 percentage points for DivPrune and 2.05-4.20 percentage points for ApET, with every 95% interval above zero. Repeating the audit on a separately resampled second manifest reproduces all six increases. In a pooled monitor audit across the four benchmarks, support Jaccard instability fails the practical AUROC criterion of 0.65 in all six settings, whereas clean compressed-model uncertainty clears it. Thus, support turnover is a weak ranking signal under the audited benchmark mixture, while clean compressed-model uncertainty is substantially more informative. These results suggest that future visual-token compression evaluations should report conditional reliability alongside clean accuracy and efficiency.

0.32

Certification of Real Images through Calibrated Content Authentication

cs.CV ·2026-10-05 ·Sarim Hashmi, Abdelrahman Elsayed, Mohammed Talha Alam, Samuele Poppi, Nils Lukas
position similarity 0.59
Abstract

Generative models can synthesize high-quality inauthentic multimedia content that is already being misused at scale. We evaluate twenty deepfake detectors against ten generators released in the last four years and find accuracy decreasing over time, from near-perfect 99.5% to 76%. Adversarial perturbations further reduce every baseline detector to below 2% accuracy, effectively inverting the detector's assigned label. We argue that this unreliability reflects a fundamental ambiguity: generators can reproduce authentic content exactly (e.g., through memorization), so content alone cannot reveal the true provenance label. For this reason, content produced by a generator must admit a faithful reconstruction by that same generator, and finding such a reconstruction makes synthetic provenance plausible and authenticity plausibly deniable. We therefore propose and evaluate a detection paradigm that outputs a calibrated prediction of whether authenticity is plausibly deniable: a faithful reconstruction by any known generator establishes plausible deniability, while calibration bounds how often content from known generators fails to be reproduced. Our evaluation shows that (i) our detector can be calibrated so that at most 1% of generated content is wrongly certified, an operating point at which most baseline detectors reach near-zero recall, including the strongest with 93% accuracy; (ii) calibrating a stricter security threshold on attacked samples preserves this bound against adaptive adversaries within the evaluated bounded-perturbation attack space, whose perturbations break every baseline, but does not cover arbitrary adversarial transformations; and (iii) post-hoc verifiability is eroding, as 1,116 of 3,000 Reddit images resist reproduction by a 2022 generator, but only 55 to 79 resist reproduction by 2024 generators.

0.21

After a Decade: Bringing Shadow Removal into the Real World with Agentic Training Data

cs.CV cs.AI ·2026-09-29 ·Shilin Hu, Jingyi Xu, Dimitris Samaras, Hieu Le
dataset similarity 0.59
Abstract

Shadow removal looks nearly solved on established benchmarks, yet remains brittle in the real world. Models have advanced; the paired training data they rely on have barely changed in nearly a decade. The reason is simple: obtaining a shadow-free target requires removing the occluder while keeping the scene, camera, and illumination otherwise unchanged, making diverse paired data difficult to capture. Meanwhile, large shadow detection datasets already contain diverse real-world images and masks, but no shadow-free targets. To turn this abundant but incomplete data into paired supervision, we propose an offline agentic workflow combining physics-motivated generation, failure detection, feedback-driven retry, candidate selection, and deterministic correction. Using this workflow, we construct AgenticShadow, a dataset of 17,138 image-mask-target triplets spanning general scenes, faces, and remote sensing. Our construction workflow reduces Color Distribution Difference by 50.5% over previous shadow removal work, while training existing shadow removal models on AgenticShadow reduces cross-domain LAB RMSE by 19.7-37.5%.

0.22

GuardPaint:SpeculativeSafetyDecodingforText-to-ImageGeneration

cs.CV cs.AI ·2026-08-22 ·Shreyash Dhoot, Paras Dhiman, Arsh Abbas Naqvi, Aranbi Dutta, Aman Chadha +2 more
similarity 0.59
Abstract

Text-to-image (T2I) diffusion models offer powerful visual generation, but their controllability creates a critical safety challenge: adversarial prompts can steer the denoising trajectory toward policy-violating content such as explicit nudity or graphic violence. Existing safeguards mostly act before generation through prompt filtering or after generation through image classification, leaving the diffusion process itself unguarded and often yielding only refusal rather than safe visual repair. We introduce GuardPaint, a speculative decoding framework for safe T2I generation that intervenes inside the diffusion trajectory without modifying the base model. A lightweight auditor monitors intermediate images, localizes unsafe regions, and triggers surgical inpainting repair only where needed. Candidate repairs are generated by a policy-aligned inpainter and selected through a guarded tournament that accepts edits only when they improve policy compliance while preserving prompt fidelity and perceptual quality. Across five jailbreak families SneakPrompt, MMA, PGJ, DACA, and RABell and UNet/flow-matching models including SD~1.5, SDXL, SD~3.5, and FLUX.1-dev. GuardPaint reduces attack success and harmful generations with minimal degradation to image quality, prompt fidelity, and benign behavior. Content warning: This paper contains examples involving nudity and violence that some readers may find disturbing, distressing, or offensive.

0.14

Shaping Useful Noise: Energy Distributions Predict Visual Pretraining Quality

cs.LG cs.AIcs.CV ·2026-11-09 ·Ching Lam Choi, Antonio Torralba, Phillip Isola, Stefanie Jegelka
top venue similarity 0.59
Abstract · Accepted as an oral presentation to NeurIPS 2026 (Sydney)

Synthetic data can scale visual pretraining, but current selection criteria only partially explain which datasets train strong representations. We propose using the energy profile of a dataset--the distribution of log-probability levels it occupies under a normalised real-image density model--as an audit and curation principle for synthetic data. To operationalise this view, we introduce Energy-Level Distribution Distance (ELD), which compares real and synthetic energy distributions. Real images span sparse high-likelihood scenes and dense low-likelihood textures, so useful synthetic data should cover this energy range rather than collapse to likely but low-variation samples. Across diverse synthetic datasets, ELD exposes mismatches missed by FID and predicts pretraining performance beyond FID, LPIPS variation, and feature-space recall, including across CNN and ViT self-supervised settings. We then use ELD to resample and mix synthetic sources, matching the real-image energy histogram while preserving diversity and improving downstream transfer.

0.32