Benchmark Scenario

In-The-Wild Robustness

Open-domain media speech. The scenario score aggregates its trial sets using the fixed OpenSVBench reporting protocol.

25 ranked systems 4 trial sets 2.3860 leading EER

Scenario Ranking

Systems on In-The-Wild Robustness

Evidence cells list trial-set rank and EER in this order: CN-Celeb · VoxCeleb1-O · VoxCeleb1-E · VoxCeleb1-H.

Rank	System	Submitted by	Scenario EER ↓	Scenario minDCF ↓	Trial-Set Evidence Rank / EER
#1	w2v-bert2.0 + cc + ecapa-tdnn + astp + sf2 w2v-bert2.0 + cc + ecapa-tdnn + astp + sf2	JinghanPeng	2.3860	0.102132	#2 4.4044 #2 0.1542 #2 0.3046 #2 0.6443
#2	W2V-BERT 2.0 + Adapter-MFA (cross-device consistency FT + weight soup) ActionPower VoicePrint	ActionPower	3.3339	0.136814	#5 6.1954 #4 0.2286 #5 0.3842 #6 0.8044
#3	MECT-b2-czur-finetuned MECT-b2-czur-finetuned	czur	3.4420	0.154836	#4 5.6266 #18 0.9095 #18 1.1318 #16 1.7309
#4	AntSpeaker/MECT-B2 antspeaker-mect-b2-vb2-full-core	Public checkpoint	3.7217	0.130256	#7 7.0910 #5 0.2287 #1 0.2901 #1 0.5387
#5	W2V-BERT 2.0 + MHFA (VoxCeleb2 + VoxBlink2) Wav2Vec2-BERT	pengjy	3.7710	0.155142	#6 7.0671 #7 0.2658 #6 0.4005 #5 0.7583
#6	palabraai/redimnet2-b6-vb2+vox2 redimnet2-b6-vb2vox2-lm-full-core	Public checkpoint	4.1123	0.141370	#9 7.8288 #3 0.1701 #4 0.3387 #3 0.6788
#7	Dual-LoRA W2V-BERT2-Adapter-MFA (TidyVoice) duallora-w2vbert2-tidyvoice-full-core	sgqt	4.2581	0.150654	#10 7.8964 #9 0.3776 #10 0.5436 #8 0.9380
#8	Agora Voiceprint (ResNet293) agora_vp	agora	4.8843	0.165519	#11 8.9045 #12 0.6170 #11 0.6878 #11 1.2872
#9	wespeaker/w2vbert2 vb2-w2vbert	Public checkpoint	5.1834	0.162046	#12 9.9690 #1 0.1489 #3 0.3146 #4 0.7298
#10	wespeaker/samresnet100 vb2-sam100	Public checkpoint	5.2487	0.170274	#13 9.9803 #6 0.2287 #8 0.4564 #7 0.8665
#11	iic/eres2netv2-zh eresv2_zh	Public checkpoint	5.5369	0.294964	#1 3.8243 #22 6.3852 #22 6.3659 #22 8.9975
#12	wespeaker/res152-voxceleb wespeaker_resnet152	Public checkpoint	6.1011	0.224787	#14 11.3672 #10 0.4998 #13 0.7174 #12 1.2879
#13	palabraai/redimnet2-b6 redimnet2	Public checkpoint	6.4216	0.208137	#17 12.2444 #8 0.2871 #9 0.5198 #10 0.9892
#14	wespeaker/res34-voxceleb wespeaker_resnet34_voxceleb	Public checkpoint	6.5576	0.240625	#15 11.9718 #15 0.8188 #15 0.9347 #14 1.6768
#15	wespeaker/res293-voxceleb wespeaker_resnet293	Public checkpoint	6.6344	0.231795	#18 12.4334 #11 0.5159 #12 0.7016 #13 1.2890
#16	iic/eres2net-en eres_en	Public checkpoint	6.6673	0.251235	#16 12.1487 #14 0.8028 #16 0.9682 #17 1.7866
#17	iic/campplus campplus	Public checkpoint	6.8211	0.361956	#3 4.7029 #24 7.7917 #24 7.9032 #24 11.1228
#18	wespeaker/res34-cnceleb wespeaker_r34	Public checkpoint	7.8434	0.404734	#8 7.4729 #23 7.0897 #23 7.1685 #23 10.3836
#19	speechbrain/ecapa ecapa	Public checkpoint	7.9977	0.317381	#20 15.2177 #17 0.9038 #7 0.4563 #9 0.9731
#20	wespeaker/campplus-voxceleb campplus_wespeaker	Public checkpoint	8.4036	0.311533	#21 15.7139 #13 0.7180 #14 0.8789 #15 1.6832
#21	wespeaker/ecapa1024-voxceleb wespeaker_ecapa1024	Public checkpoint	8.5204	0.308148	#22 15.8209 #16 0.8188 #17 0.9899 #18 1.8513
#22	wespeaker/ecapa512-voxceleb wespeaker_ecapa512	Public checkpoint	8.9203	0.324306	#23 16.3391 #19 1.0371 #19 1.2104 #19 2.2570
#23	iic/eres2net-large-3dspeaker eres2net_large_3dspeaker	Public checkpoint	14.2563	0.581636	#19 14.0524 #25 13.0837 #25 13.1915 #25 17.1053
#24	microsoft/wavlm-base-plus-sv wavlm_base	Public checkpoint	16.5057	0.632600	#24 27.9358 #20 5.5047 #20 3.2098 #20 6.5121
#25	microsoft/unispeech-sat-base-plus-sv unispeech_sat_base_plus	Public checkpoint	17.6846	0.670598	#25 29.8022 #21 5.6037 #21 3.6382 #21 7.4593