Results - Media Analytics - ELSA Benchmarks Platform

method: ViT_B_16_Weights2024-05-31

Authors: Witold Pietroń, Aleksander Samek

Affiliation: Warsaw University of Technology

Description: We used pretrained model ViT_B_16_Weights pretrained model and trained it on provided dataset.

method: First example2024-03-30

Authors: Lorenzo Baraldi

Description: First example

First example

method: ViT_L_16_Weights SWAG2024-06-04

Authors: Witold Pietroń, Aleksander Samek

Affiliation: Warsaw University of Technology

Description: We used pretrained model ViT_L_16_Weights pretrained model and trained it on provided dataset. Starting from ViT_L_16_Weights.IMAGENET1K_SWAG_LINEAR_V1 weights.

Ranking Table

Description Paper Source Code

Date	Method	precision	recall	f1	accuracy_gan	accuracy_diffusion	GEN_0	GEN_1	GEN_2	GEN_3	GEN_4	GEN_5	GEN_6	GEN_7	GEN_8	GEN_9	GEN_10	GEN_11	GEN_12	GEN_13	GEN_14	GEN_15	GEN_16	GEN_17	GEN_18	GEN_19	GEN_20	GEN_21	GEN_22	GEN_23	GEN_24	GEN_25	GEN_26	GEN_27	GEN_28	GEN_29	GEN_30	GEN_31	GEN_32	GEN_33	GEN_34	GEN_35	GEN_36	GEN_37	GEN_38	GEN_39	GEN_40	GEN_41	GEN_42	GEN_43
2024-05-31	ViT_B_16_Weights	0.9894	0.5263	0.6871	0.4963	0.7684	0.8409	0.5594	0.5278	0.1341	0.5742	0.8686	0.5442	0.0163	0.5755	0.5617	0.8473	0.4254	0.6950	0.1295	0.0504	0.2063	0.8923	0.0590	0.7042	0.8437	0.5703	0.4381	0.8910	0.5561	0.4454	0.4414	0.5421	0.8582	0.5888	0.5876	0.1390	0.8800	0.8842	0.0575	0.7112	0.5846	0.1280	0.5426	0.8796	0.2053	0.7242	0.0152	0.5534	0.0697
2024-03-30	First example	0.9508	0.5018	0.6569	0.5009	0.5009	0.5023	0.5083	0.4942	0.5088	0.4898	0.5127	0.5025	0.4847	0.5071	0.4923	0.4950	0.5127	0.5025	0.5013	0.5070	0.5005	0.5021	0.4974	0.5013	0.4935	0.4942	0.5035	0.4990	0.5042	0.4771	0.4931	0.5090	0.4942	0.5144	0.5077	0.5099	0.5000	0.5017	0.4944	0.5083	0.5008	0.5175	0.4958	0.5075	0.4995	0.4983	0.4962	0.5129	0.4981
2024-06-04	ViT_L_16_Weights SWAG	0.9834	0.4945	0.6581	0.5215	0.7066	0.8255	0.7470	0.4321	0.2500	0.3809	0.8258	0.3660	0.1415	0.3965	0.3637	0.8627	0.2496	0.6367	0.2181	0.1811	0.2576	0.8376	0.1710	0.5470	0.7815	0.5700	0.5604	0.8415	0.3619	0.5750	0.2411	0.3644	0.7847	0.3934	0.4109	0.2618	0.8868	0.8434	0.1898	0.5574	0.5611	0.2260	0.4158	0.8614	0.2526	0.6610	0.1206	0.7461	0.1720
2024-06-17	Code Knn	0.9921	0.6981	0.8195	0.5330	0.8696	0.9933	0.9746	0.7850	0.0766	0.8746	0.9923	0.7575	0.0640	0.9108	0.8140	0.9721	0.7506	0.9015	0.1863	0.0952	0.0800	0.9533	0.0977	0.9506	0.9952	0.9117	0.5788	0.8325	0.7752	0.5775	0.6558	0.6994	0.9456	0.8069	0.8215	0.2295	0.9156	0.9385	0.2214	0.8571	0.8225	0.2893	0.7298	0.9438	0.3005	0.8500	0.1646	0.8842	0.2710
2024-05-31	Convolutional neural network	0.9767	0.4859	0.6490	0.5440	0.6555	0.7713	0.7541	0.3333	0.3270	0.2758	0.7797	0.2781	0.2671	0.2747	0.1885	0.8906	0.3120	0.5125	0.3421	0.2925	0.3602	0.7780	0.2890	0.6319	0.7427	0.5465	0.6762	0.7766	0.1800	0.6847	0.3150	0.2608	0.7544	0.2891	0.2778	0.3386	0.9064	0.7880	0.2986	0.6536	0.5473	0.3244	0.3126	0.7960	0.3618	0.5275	0.2696	0.7529	0.2576
2024-06-01	Swin DCT with Multicrop	0.9962	0.6698	0.8011	0.5590	0.8804	0.9998	0.9602	0.9415	0.0689	0.9742	0.9998	0.9369	0.0160	0.9527	0.9690	0.8031	0.7823	0.9406	0.4003	0.3258	0.2800	0.9427	0.1824	0.9608	0.9998	0.9429	0.6777	0.9588	0.7852	0.4650	0.5079	0.7235	0.8133	0.8081	0.7744	0.0421	0.6048	0.8119	0.1786	0.7452	0.6496	0.2193	0.6746	0.8052	0.1525	0.6817	0.0295	0.6119	0.1120
2024-06-17	Code SVM	0.9864	0.8103	0.8897	0.5805	0.8577	0.9098	0.9719	0.9221	0.2354	0.9746	0.9475	0.9217	0.2501	0.9619	0.9442	0.9296	0.9485	0.9692	0.3250	0.1694	0.1725	0.9185	0.1552	0.9856	0.9506	0.9696	0.8215	0.6521	0.9190	0.8215	0.8760	0.8806	0.9494	0.9448	0.9240	0.5704	0.9290	0.9446	0.5104	0.9567	0.9275	0.5963	0.8833	0.9198	0.5760	0.9444	0.4552	0.9533	0.4913
2024-05-22	Swin Base Huge data aug	0.9956	0.6665	0.7985	0.6354	0.8506	0.9971	0.4529	0.7842	0.2801	0.8719	0.9952	0.7094	0.0990	0.8015	0.9150	0.8750	0.7367	0.9750	0.6325	0.4898	0.5095	0.9492	0.3210	0.8288	0.9956	0.5238	0.6454	0.9373	0.7623	0.5529	0.6046	0.6619	0.8652	0.7802	0.7179	0.1886	0.7660	0.8785	0.3318	0.7460	0.4798	0.4080	0.6771	0.8940	0.3065	0.7735	0.1106	0.4742	0.2536
2024-05-29	Swin Base Huge Data Aug + DCT + Random Fake (One Epoch)	0.9956	0.6707	0.8015	0.6365	0.8532	0.9940	0.5279	0.7842	0.3240	0.9090	0.9892	0.7246	0.1286	0.8417	0.9185	0.8146	0.7435	0.9788	0.5975	0.4710	0.5405	0.9444	0.3073	0.8710	0.9938	0.4931	0.6615	0.9425	0.7702	0.5438	0.6025	0.6673	0.8656	0.7994	0.7467	0.2078	0.7135	0.8746	0.3112	0.7633	0.4423	0.3820	0.6738	0.8802	0.3165	0.7956	0.1341	0.5027	0.2339
2024-05-30	Swin Base Huge Data Aug + DCT + Random Fake (Two Epochs - TH adjusted)	0.9945	0.7158	0.8324	0.6387	0.8700	0.9990	0.4958	0.8604	0.3258	0.9356	0.9975	0.7977	0.2101	0.8738	0.9558	0.9160	0.7967	0.9823	0.5628	0.4262	0.5055	0.9340	0.3399	0.8769	0.9988	0.6756	0.7013	0.9129	0.8331	0.6358	0.6873	0.7488	0.9152	0.8492	0.7913	0.2354	0.8229	0.9196	0.3460	0.8025	0.6100	0.4148	0.7617	0.9240	0.3570	0.8206	0.2171	0.5110	0.3081
2024-05-29	Swin Base Huge Data Aug + DCT + Random Fake (One Epoch - TH adjusted)	0.9905	0.8077	0.8898	0.7032	0.8649	0.9992	0.7269	0.9069	0.5043	0.9673	0.9985	0.8744	0.4427	0.9308	0.9654	0.9248	0.9140	0.9948	0.8020	0.7072	0.7825	0.8602	0.5965	0.9627	0.9990	0.7800	0.7856	0.8408	0.8775	0.7223	0.7952	0.8304	0.9319	0.8983	0.8646	0.3600	0.8602	0.9346	0.5100	0.8913	0.6921	0.5940	0.8265	0.9444	0.5320	0.8873	0.3727	0.6906	0.4678
2024-05-24	PatchNet2	0.9924	0.6128	0.7578	0.7229	0.7744	0.9715	0.8777	0.7096	0.7274	0.8523	0.9817	0.7506	1.0000	0.8073	0.5592	0.8979	0.2915	0.5404	0.9378	0.9836	0.9540	0.9106	0.8917	0.7354	0.9560	0.7688	0.2485	0.9096	0.4575	0.2927	0.2842	0.5760	0.7131	0.6288	0.5894	0.1372	0.6131	0.7373	0.1308	0.5629	0.5148	0.1610	0.5131	0.7127	0.1420	0.4031	0.2526	0.6146	0.1113
2024-06-07	CLIP-adapter	0.9932	0.8522	0.9173	0.7612	0.8987	1.0000	0.8971	0.9294	0.3742	0.9725	1.0000	0.9938	0.1696	0.9915	0.9998	0.9300	0.8246	1.0000	0.9390	0.8372	0.7105	0.8898	0.5079	0.9913	1.0000	0.9542	0.8108	0.8848	0.9283	0.7475	0.7010	0.8998	0.9400	0.9050	0.9144	0.5151	0.8504	0.9475	0.7368	0.8992	0.8108	0.8425	0.8254	0.9531	0.7225	0.9113	0.5903	0.7752	0.6745
2024-05-30	UniFID - K_means	0.9907	0.8263	0.9010	0.8510	0.8395	0.8304	0.7788	0.6969	0.5859	0.7727	0.8317	0.6804	0.7809	0.8096	0.6440	0.7646	0.7252	0.8288	0.7983	0.8108	0.8080	0.7117	0.8017	0.7727	0.8254	0.7215	0.5125	0.9888	0.7021	0.6465	0.9496	0.8390	1.0000	0.9594	0.9785	0.6595	0.8956	0.9998	0.9996	0.9815	0.8675	0.9920	0.8500	1.0000	0.9920	0.9996	0.9995	1.0000	0.9924
2024-05-31	UniFID - V_exp3	0.9969	0.8210	0.9004	0.8649	0.8971	0.8323	0.7735	0.7652	0.4532	0.8108	0.8263	0.7894	0.6433	0.8285	0.7008	0.8238	0.7575	0.7225	0.6465	0.6508	0.6545	0.9142	0.6563	0.8040	0.8271	0.7338	0.6150	0.9888	0.7021	0.6465	0.9496	0.8390	1.0000	0.9594	0.9785	0.6595	0.8956	0.9998	0.9996	0.9815	0.8675	0.9920	0.8500	1.0000	0.9920	0.9996	0.9995	1.0000	0.9924
2024-06-01	UniFID - exp5_Method2	0.9980	0.7958	0.8855	0.8773	0.8892	0.8067	0.7496	0.6504	0.4575	0.7385	0.8073	0.6369	0.6568	0.7763	0.5917	0.7277	0.6852	0.7425	0.6580	0.6716	0.6680	0.9508	0.6715	0.7358	0.7983	0.6817	0.4610	0.9888	0.7021	0.6465	0.9496	0.8390	1.0000	0.9594	0.9785	0.6595	0.8956	0.9998	0.9996	0.9815	0.8675	0.9920	0.8500	1.0000	0.9920	0.9996	0.9995	1.0000	0.9924
2024-05-28	UniFID - tow_modles_threshold	0.9979	0.8008	0.8886	0.8774	0.8910	0.8131	0.7590	0.6667	0.4617	0.7496	0.8135	0.6515	0.6648	0.7850	0.6113	0.7350	0.6983	0.7479	0.6630	0.6732	0.6735	0.9465	0.6752	0.7498	0.8052	0.6981	0.4815	0.9888	0.7021	0.6465	0.9496	0.8390	1.0000	0.9594	0.9785	0.6595	0.8956	0.9998	0.9996	0.9815	0.8675	0.9920	0.8500	1.0000	0.9920	0.9996	0.9995	1.0000	0.9924
2024-05-31	UniFID - cont_thre84.6	0.9979	0.8008	0.8886	0.8774	0.8910	0.8131	0.7590	0.6667	0.4617	0.7496	0.8135	0.6515	0.6648	0.7850	0.6113	0.7350	0.6983	0.7479	0.6630	0.6732	0.6735	0.9465	0.6752	0.7498	0.8052	0.6981	0.4815	0.9888	0.7021	0.6465	0.9496	0.8390	1.0000	0.9594	0.9785	0.6595	0.8956	0.9998	0.9996	0.9815	0.8675	0.9920	0.8500	1.0000	0.9920	0.9996	0.9995	1.0000	0.9924
2024-05-31	Deepfake Detection by ConvNext	0.9960	0.8871	0.9384	0.8807	0.9134	1.0000	0.9992	0.9883	1.0000	0.9971	1.0000	0.9629	0.9955	0.9981	0.9940	0.9971	0.9990	0.8175	0.9998	1.0000	1.0000	0.9448	0.9281	0.9942	1.0000	0.9752	0.9973	0.9179	0.8675	0.8263	0.8529	0.7777	0.8508	0.8719	0.8750	0.7398	0.7863	0.8290	0.7714	0.8402	0.7756	0.7753	0.7952	0.8500	0.6665	0.5400	0.6778	0.8071	0.4058

Media Analytics

Inactive evaluations

method: ViT_B_16_Weights2024-05-31

method: First example2024-03-30

method: ViT_L_16_Weights SWAG2024-06-04

Ranking Table

Ranking Graphic

Ranking Graphic