SOTAVerified

Image Attribution

Image attribution algorithms aim to identify important regions that are highly relevant to model decisions.

Papers

Showing 1–26 of 26 papers

TitleStatusHype
What the DAAM: Interpreting Stable Diffusion Using Cross AttentionCode2
Less is More: Fewer Interpretable Region via Submodular Subset SelectionCode2
Axiomatic Attribution for Deep NetworksCode1
Making Sense of Dependence: Efficient Black-box Explanations Using Dependence MeasureCode1
Deepfake Network Architecture AttributionCode1
"Why Should I Trust You?": Explaining the Predictions of Any ClassifierCode1
RepMix: Representation Mixing for Robust Attribution of Synthesized ImagesCode1
Reverse Engineering of Generative Models: Inferring Model Hyperparameters from Generated ImagesCode1
Gifsplanation via Latent Shift: A Simple Autoencoder Approach to Counterfactual Generation for Chest X-raysCode1
RISE: Randomized Input Sampling for Explanation of Black-box ModelsCode1
Grad-CAM: Visual Explanations from Deep Networks via Gradient-based LocalizationCode1
A Unified Approach to Interpreting Model PredictionsCode1
LoRAX: LoRA eXpandable Networks for Continual Synthetic Image AttributionCode0
A Siamese-based Verification System for Open-set Architecture Attribution of Synthetic ImagesCode0
Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency MapsCode0
Detecting Origin Attribution for Text-to-Image Diffusion ModelsCode0
Attributing Fake Images to GANs: Learning and Analyzing GAN FingerprintsCode0
WILD: a new in-the-Wild Image Linkage Dataset for synthetic image attribution—0
ARIA: Adversarially Robust Image Attribution for Content Provenance—0
BOSC: A Backdoor-based Framework for Open Set Synthetic Image Attribution—0
Hide and Seek: How Does Watermarking Impact Face Recognition?—0
I2AM: Interpreting Image-to-Image Latent Diffusion Models via Attribution Maps—0
Learning to Disentangle GAN Fingerprint for Fake Image Attribution—0
OSCAR-Net: Object-centric Scene Graph Attention for Image Attribution—0
Sampling Matters in Explanations: Towards Trustworthy Attribution Analysis Building Block in Visual Models through Maximizing Explanation Certainty—0
Are handcrafted filters helpful for attributing AI-generated images?—0
Show:102550

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1SMDL-Attribution (ICLR version)Insertion AUC score (ArcFace ResNet-101)0.58—Unverified
2RISEInsertion AUC score (ArcFace ResNet-101)0.57—Unverified
3HSIC-AttributionInsertion AUC score (ArcFace ResNet-101)0.57—Unverified
4LIMEInsertion AUC score (ArcFace ResNet-101)0.52—Unverified
5Kernel SHAPInsertion AUC score (ArcFace ResNet-101)0.52—Unverified
6SaliencyInsertion AUC score (ArcFace ResNet-101)0.46—Unverified
7Grad-CAMInsertion AUC score (ArcFace ResNet-101)0.37—Unverified
8Integrated GradientsInsertion AUC score (ArcFace ResNet-101)0.36—Unverified
#ModelMetricClaimedVerifiedStatus
1SMDL-Attribution (ICLR version)Insertion AUC score (ResNet-101)0.73—Unverified
2Grad-CAMInsertion AUC score (ResNet-101)0.72—Unverified
3RISEInsertion AUC score (ResNet-101)0.72—Unverified
4HSIC-AttributionInsertion AUC score (ResNet-101)0.68—Unverified
5LIMEInsertion AUC score (ResNet-101)0.68—Unverified
6Kernel SHAPInsertion AUC score (ResNet-101)0.68—Unverified
7SaliencyInsertion AUC score (ResNet-101)0.66—Unverified
8Integrated GradientsInsertion AUC score (ResNet-101)0.04—Unverified
#ModelMetricClaimedVerifiedStatus
1SMDL-Attribution (ICLR version)Insertion AUC score (ArcFace ResNet-101)0.67—Unverified
2HSIC-AttributionInsertion AUC score (ArcFace ResNet-101)0.67—Unverified
3RISEInsertion AUC score (ArcFace ResNet-101)0.65—Unverified
4LIMEInsertion AUC score (ArcFace ResNet-101)0.62—Unverified
5Kernel SHAPInsertion AUC score (ArcFace ResNet-101)0.61—Unverified
6SaliencyInsertion AUC score (ArcFace ResNet-101)0.56—Unverified
7Integrated GradientsInsertion AUC score (ArcFace ResNet-101)0.54—Unverified
8Grad-CAMInsertion AUC score (ArcFace ResNet-101)0.47—Unverified