This table lists the benchmark results for the Binary Classification scenario.


Method InfoDeepfakesFace2FaceFaceSwapNeuralTexturesPristineTotal
sorted bysort bysort bysort bysort bysort by
StableForensics0.9910.8470.9510.7870.8840.883
Two-stream-SRM-RGB10.9820.9270.9420.9530.1740.562
PredictFake0.9730.8470.9130.8200.8940.887
RobustForensics0.9730.8320.9420.7600.8540.859
eff-b7-v30.9730.9120.9130.8070.1980.546
Sentinel0.9640.9050.8830.8670.6240.763
face single model0.9640.8690.9420.4600.7380.760
Xceptionpermissive0.9640.8690.9030.8070.5240.710
Andreas Rössler, Davide Cozzolino, Luisa Verdoliva, Christian Riess, Justus Thies, Matthias Nießner: FaceForensics++: Learning to Detect Manipulated Facial Images. ICCV 2019
LGSC_Lite0.9550.8390.8740.7730.8440.848
fakeface0.9550.7660.8830.7470.8060.816
eff-b7-att epc110.9550.7880.8640.5470.5920.680
EfficientNet-b7-Attention0.9550.8030.8540.6000.6160.701
yd dust face0.9550.8610.9510.6930.7620.806
simple policybinary0.9550.8690.9420.3870.7440.751
EfficientNet-b40.9550.7960.8250.8270.7120.779
Firefly0.9550.7300.8740.6670.9200.855
ATDETECTOR0.9550.7960.9220.7800.8980.875
RealFace0.9450.7660.8640.8130.7900.815
faceClassify10.9450.8540.9810.7930.7300.806
ZFake0.9450.7370.8640.6930.8800.838
Inception Resnet V1permissive0.9360.8390.9030.8200.7500.809
Nika Dogonadze, Jana Obernosterer: Deep Face Forgery Detection. Advanced Deep Learning for Computer Vision Course at TUM
antifake0.9360.8030.9420.6870.7640.795
UltraVision0.9180.8470.9320.7330.8100.828
swap_classify0.9090.7590.8640.7470.7240.767
SCANcopyleft0.9090.8250.7380.6130.7640.763
IR+unet0.9090.8470.8450.7870.7900.816
unet+res0.8820.8320.7960.7670.8120.814
EfficientB4_3800.8820.7010.7960.7400.7940.783
MesoNet0.8730.5620.6120.4070.7260.660
Darius Afchar, Vincent Nozick, Junichi Yamagishi, and Isao Echizen: Mesonet: a compact facial video forgery detection network. arXiv
framelevel0.8730.7370.8350.7600.7600.777
LVLNet0.8640.5260.6990.2470.8820.717
Recasting0.8550.6790.7380.7800.3440.552
Davide Cozzolino, Giovanni Poggi, and Luisa Verdoliva: Recasting residual-based local descriptors as convolutional neural networks: an application to image forgery detection. ACM Workshop on Information Hiding and Multimedia Security
Rahmouni0.8550.6420.5630.6070.5000.581
Nicolas Rahmouni, Vincent Nozick, Junichi Yamagishi, and Isao Echizen: Distinguishing computer graphics from natural images using convolution neural networks. IEEE Workshop on Information Forensics and Security,
Bayar and Stamm0.8450.7370.8250.7070.4620.616
Belhassen Bayar and Matthew C. Stamm: A deep learning approach to universal image manipulation detection using a new convolutional layer. ACM Workshop on Information Hiding and Multimedia Security
EfficientNetB70.8360.8470.8740.7470.6300.725
inceptionv10.8270.7300.8060.5330.4200.564
InceptResV10.8270.7300.8060.5330.4200.564
HRC0.8270.6130.5730.6600.8480.757
Forged face detection : fCNN0.7910.6420.7090.5130.5440.597
framelevel_all_compress_format0.7820.1240.1550.1730.8600.575
XceptionNet Full Imagepermissive0.7450.7590.7090.7330.5100.624
Andreas Rössler, Davide Cozzolino, Luisa Verdoliva, Christian Riess, Justus Thies, Matthias Nießner: FaceForensics++: Learning to Detect Manipulated Facial Images. ICCV 2019
Steganalysis Features0.7360.7370.6890.6330.3400.518
Jessica Fridrich and Jan Kodovsky: Rich Models for Steganalysis of Digital Images. IEEE Transactions on Information Forensics and Security
GAEL-Net0.7180.6860.6310.7070.5620.625
Jae-Yong Baek, Yong-Sang Yoo, Seung-Hwan Bae: Generative Adversarial Ensemble Learning for Face Forensics. IEEE Access
YSNet0.5640.5840.5920.4270.4920.513
resnet500.5270.5040.4080.3870.7860.620
IAMFAKE20.4640.4960.5150.3670.8260.640
IAMFAKE0.4360.4090.5240.3000.8160.611