EURASIP Journal on Information Security

Table 4 Theoretical performance of the individual detectors with the combined LondonDB/UtrechtDB/Alabama datasets (best result per morph type marked in bold)

From: Potential advantages and limitations of using information fusion in media forensics—a discussion on the example of detecting face morphing attacks

		*SC1*		*SC2*
Detector	Morph type	EER	τ_adaptive	EER	τ_adaptive
D_ArXivMC	Combined	3.95%	0.528241	8.27%	0.417467
D_ArXivNaive		1.94%	0.594687	4.96%	0.499938
D_BIOSIGMC		9.75%	0.617098	16.31%	0.561516
D_BIOSIGNaive		7.74%	0.558175	13.56%	0.478364
D_keypoints		12.65%	0.971509	19.08%	0.990942
D_ArXivMC	Complete	4.00%	0.526729	7.75%	0.424468
D_ArXivNaive		1.82%	0.600357	4.06%	0.507648
D_BIOSIGMC		9.75%	0.616997	14.98%	0.565297
D_BIOSIGNaive		7.45%	0.563476	12.07%	0.496052
keypoints		12.43%	0.972011	19.53%	0.990758
D_ArXivMC	Splicing	4.99%	0.501098	9.37%	0.406828
D_ArXivNaive		2.76%	0.566983	5.37%	0.492199
D_BIOSIGMC		12.67%	0.594189	20.64%	0.541497
D_BIOSIGNaive		9.08%	0.54235	14.84%	0.470685
D_keypoints		11.09%	0.976876	19.08%	0.990933

Back to article page