SOTAVerified

Image Manipulation Localization

The task of segmenting parts of images or image parts that have been tampered with or manipulated (sometimes also referred to as doctored). This typically encompasses image splicing, copy-move, or image inpainting.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–31 of 31 papers

TitleStatusHype
Towards Modern Image Manipulation Localization: A Large-Scale Dataset and Novel MethodsCode3
IMDL-BenCo: A Comprehensive Benchmark and Codebase for Image Manipulation Detection & LocalizationCode3
Can We Get Rid of Handcrafted Feature Extractors? SparseViT: Nonsemantics-Centered, Parameter-Efficient Image Manipulation Localization through Spare-Coding TransformerCode2
IML-ViT: Benchmarking Image Manipulation Localization by Vision TransformerCode2
CMX: Cross-Modal Fusion for RGB-X Semantic Segmentation with TransformersCode2
Mesoscopic Insights: Orchestrating Multi-scale & Hybrid Architecture for Image Manipulation LocalizationCode2
ForensicHub: A Unified Benchmark & Codebase for All-Domain Fake Image Detection and LocalizationCode2
Learning JPEG Compression Artifacts for Image Manipulation Detection and LocalizationCode1
MVSS-Net: Multi-View Multi-Scale Supervised Networks for Image Manipulation DetectionCode1
MMFusion: Combining Image Forensic Filters for Visual Manipulation Detection and LocalizationCode1
Image Manipulation Detection by Multi-View Multi-Scale SupervisionCode1
Pre-training-free Image Manipulation Localization through Non-Mutually Exclusive Contrastive LearningCode1
PSCC-Net: Progressive Spatio-Channel Correlation Network for Image Manipulation Detection and LocalizationCode1
SPAN: Spatial Pyramid Attention Network for Image Manipulation LocalizationCode1
TruFor: Leveraging all-round clues for trustworthy image forgery detection and localization—0
A Noise and Edge extraction-based dual-branch method for Shallowfake and Deepfake Localization—0
Beyond Fully Supervised Pixel Annotations: Scribble-Driven Weakly-Supervised Framework for Image Manipulation Localization—0
Constrained R-CNN: A general image manipulation detection model—0
Context-Aware Weakly Supervised Image Manipulation Localization with SAM Refinement—0
ForgeryTTT: Zero-Shot Image Manipulation Localization with Test-Time Training—0
M^3:Manipulation Mask Manufacturer for Arbitrary-Scale Super-Resolution Mask—0
ObjectFormer for Image Manipulation Detection and Localization—0
Omni-IML: Towards Unified Image Manipulation Localization—0
Perceptual MAE for Image Manipulation Localization: A High-level Vision Learner Focusing on Low-level Features—0
Pixel-Inconsistency Modeling for Image Manipulation Localization—0
PROMPT-IML: Image Manipulation Localization with Pre-trained Foundation Models Through Prompt Tuning—0
SPN-CNN: Boosting Sensor-Based Source Camera Attribution With Deep Learning—0
TBNet:Two-Stream Boundary-aware Network for Generic Image Manipulation Localization—0
TriPINet: Tripartite Progressive Integration Network for Image Manipulation Localization—0
Digital Image Forensics: A quantitative & qualitative comparison between State-of-the-art-AI and Traditional Techniques for detection and localization of image manipulationsCode0
ManTra-Net: Manipulation Tracing Network for Detection and Localization of Image Forgeries With Anomalous FeaturesCode0
Show:102550
All datasetsCasia V1+CocoGlideColumbiaCOVERAGEDSO-1CASIAv1(Protoclo-CAT)Columbia(Protocol-CAT)COVERAGE(Protocol-CAT)NIST16(Protocol-CAT)

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1CMX (RGB+SRM)Average Pixel F1(Fixed threshold)0.79—Unverified
2Early FusionAverage Pixel F1(Fixed threshold)0.78—Unverified
3Late FusionAverage Pixel F1(Fixed threshold)0.78—Unverified
4CMX (RGB+Bayar)Average Pixel F1(Fixed threshold)0.77—Unverified
5CMX (RGB+NP++)Average Pixel F1(Fixed threshold)0.76—Unverified
6CAT-Net v2Average Pixel F1(Fixed threshold)0.75—Unverified
7TruForAverage Pixel F1(Fixed threshold)0.74—Unverified
8MVSS-NetAverage Pixel F1(Fixed threshold)0.53—Unverified
9CR-CNNAverage Pixel F1(Fixed threshold)0.48—Unverified
10ManTraNetAverage Pixel F1(Fixed threshold)0.18—Unverified
#ModelMetricClaimedVerifiedStatus
1CMX (RGB+SRM)Average Pixel F1(Fixed threshold)0.59—Unverified
2Late FusionAverage Pixel F1(Fixed threshold)0.57—Unverified
3CMX (RGB+Bayar)Average Pixel F1(Fixed threshold)0.57—Unverified
4Early FusionAverage Pixel F1(Fixed threshold)0.55—Unverified
5TruForAverage Pixel F1(Fixed threshold)0.52—Unverified
6ManTraNetAverage Pixel F1(Fixed threshold)0.52—Unverified
7CMX (RGB+NP++)Average Pixel F1(Fixed threshold)0.52—Unverified
8MVSS-NetAverage Pixel F1(Fixed threshold)0.49—Unverified
9CR-CNNAverage Pixel F1(Fixed threshold)0.45—Unverified
10CAT-Net v2Average Pixel F1(Fixed threshold)0.43—Unverified
#ModelMetricClaimedVerifiedStatus
1Early FusionAverage Pixel F1(Fixed threshold)0.89—Unverified
2CMX (RGB+NP++)Average Pixel F1(Fixed threshold)0.88—Unverified
3CMX (RGB+Bayar)Average Pixel F1(Fixed threshold)0.87—Unverified
4Late FusionAverage Pixel F1(Fixed threshold)0.86—Unverified
5CAT-Net v2Average Pixel F1(Fixed threshold)0.86—Unverified
6TruForAverage Pixel F1(Fixed threshold)0.86—Unverified
7CMX (RGB+SRM)Average Pixel F1(Fixed threshold)0.83—Unverified
8SPANAverage Pixel F1(Fixed threshold)0.76—Unverified
9MVSS-NetAverage Pixel F1(Fixed threshold)0.73—Unverified
10CR-CNNAverage Pixel F1(Fixed threshold)0.63—Unverified
#ModelMetricClaimedVerifiedStatus
1Early FusionAverage Pixel F1(Fixed threshold)0.66—Unverified
2Late FusionAverage Pixel F1(Fixed threshold)0.64—Unverified
3CMX (RGB+SRM)Average Pixel F1(Fixed threshold)0.63—Unverified
4TruForAverage Pixel F1(Fixed threshold)0.6—Unverified
5CMX (RGB+Bayar)Average Pixel F1(Fixed threshold)0.59—Unverified
6CMX (RGB+NP++)Average Pixel F1(Fixed threshold)0.58—Unverified
7MVSS-NetAverage Pixel F1(Fixed threshold)0.51—Unverified
8CR-CNNAverage Pixel F1(Fixed threshold)0.39—Unverified
9CAT-Net v2Average Pixel F1(Fixed threshold)0.38—Unverified
10ManTraNetAverage Pixel F1(Fixed threshold)0.32—Unverified
#ModelMetricClaimedVerifiedStatus
1TruForAverage Pixel F1(Fixed threshold)0.93—Unverified
2Late FusionAverage Pixel F1(Fixed threshold)0.9—Unverified
3CMX (RGB+NP++)Average Pixel F1(Fixed threshold)0.9—Unverified
4Early FusionAverage Pixel F1(Fixed threshold)0.87—Unverified
5CMX (RGB+SRM)Average Pixel F1(Fixed threshold)0.79—Unverified
6CMX (RGB+Bayar)Average Pixel F1(Fixed threshold)0.78—Unverified
7CAT-Net v2Average Pixel F1(Fixed threshold)0.58—Unverified
8ManTraNetAverage Pixel F1(Fixed threshold)0.41—Unverified
9MVSS-NetAverage Pixel F1(Fixed threshold)0.36—Unverified
10CR-CNNAverage Pixel F1(Fixed threshold)0.29—Unverified
#ModelMetricClaimedVerifiedStatus
1MesorchPixel Binary F10.84—Unverified
2TruforPixel Binary F10.82—Unverified
3CAT-NetPixel Binary F10.81—Unverified
4IML-ViTPixel Binary F10.8—Unverified
5MVSS-NetPixel Binary F10.6—Unverified
6PSCC-NetPixel Binary F10.59—Unverified
7ObjectFormerPixel Binary F10.53—Unverified
8Mantra-NetPixel Binary F10.33—Unverified
#ModelMetricClaimedVerifiedStatus
1IML-ViTPixel Binary F10.95—Unverified
2MesorchPixel Binary F10.92—Unverified
3CAT-NetPixel Binary F10.92—Unverified
4TruforPixel Binary F10.89—Unverified
5PSCC-NetPixel Binary F10.86—Unverified
6MVSS-NetPixel Binary F10.74—Unverified
7ObjectFormerPixel Binary F10.73—Unverified
8Mantra-NetPixel Binary F10.46—Unverified
#ModelMetricClaimedVerifiedStatus
1IML-ViTPixel Binary F10.65—Unverified
2MesorchPixel Binary F10.59—Unverified
3MVSS-NetPixel Binary F10.5—Unverified
4TruforPixel Binary F10.46—Unverified
5CAT-NetPixel Binary F10.43—Unverified
6PSCCPixel Binary F10.38—Unverified
7ObjectFormerPixel Binary F10.26—Unverified
8Mantra-NetPixel Binary F10.2—Unverified
#ModelMetricClaimedVerifiedStatus
1IML-ViTPixel Binary F10.5—Unverified
2MesorchPixel Binary F10.39—Unverified
3PSCC-NetPixel Binary F10.37—Unverified
4MVSS-NetPixel Binary F10.35—Unverified
5TruforPixel Binary F10.35—Unverified
6ObjectFormerPixel Binary F10.25—Unverified
7CAT-NetPixel Binary F10.25—Unverified
8Mantra-NetPixel Binary F10.19—Unverified