Trial Set

Lombard Grid Lombard

Builds plain-to-Lombard SV trials from the public Lombard Grid release because no upstream SV trial keys are used.

Source dataset: Lombard Grid
Trials: 29524
Protocol source: Self-constructed protocol
Source split: Public Lombard Grid audio release

lombardgrid_lombard Speaking style

Source Dataset

About Lombard Grid

The Lombard Grid corpus records English speakers producing ordinary and Lombard speech, the louder, more effortful style people often adopt in noise. It includes synchronized frontal and profile video, although OpenSVBench uses only the audio. The release contains 5,400 utterances from 54 talkers, split evenly between Lombard and plain reference speech.

Language: English
Speakers: 54 talkers
Recordings: 5,400 utterances
Conditions: Plain speech and noise-induced Lombard speech

Direct Ranking

Systems on this trial set

25 reviewed systems currently have valid results.

Rank	System	Submitted by	Trial-Set EER ↓	Trial-Set minDCF ↓
#1	AntSpeaker/MECT-B2 antspeaker-mect-b2-vb2-full-core	Public checkpoint	0.0075	0.000373
#2	W2V-BERT 2.0 + MHFA (VoxCeleb2 + VoxBlink2) Wav2Vec2-BERT	pengjy	0.0149	0.001118
#3	Dual-LoRA W2V-BERT2-Adapter-MFA (TidyVoice) duallora-w2vbert2-tidyvoice-full-core	sgqt	0.0186	0.001863
#4	palabraai/redimnet2-b6-vb2+vox2 redimnet2-b6-vb2vox2-lm-full-core	Public checkpoint	0.0373	0.001453
#5	w2v-bert2.0 + cc + ecapa-tdnn + astp + sf2 w2v-bert2.0 + cc + ecapa-tdnn + astp + sf2	JinghanPeng	0.0373	0.001826
#6	wespeaker/samresnet100 vb2-sam100	Public checkpoint	0.0373	0.004657
#7	wespeaker/w2vbert2 vb2-w2vbert	Public checkpoint	0.0559	0.005104
#8	MECT-b2-czur-finetuned MECT-b2-czur-finetuned	czur	0.0596	0.003279
#9	W2V-BERT 2.0 + Adapter-MFA (cross-device consistency FT + weight soup) ActionPower VoicePrint	ActionPower	0.0633	0.005067
#10	Agora Voiceprint (ResNet293) agora_vp	agora	0.0708	0.002534
#11	wespeaker/res152-voxceleb wespeaker_resnet152	Public checkpoint	0.0745	0.011736
#12	palabraai/redimnet2-b6 redimnet2	Public checkpoint	0.0782	0.007675
#13	wespeaker/res34-voxceleb wespeaker_resnet34_voxceleb	Public checkpoint	0.1490	0.010991
#14	wespeaker/res293-voxceleb wespeaker_resnet293	Public checkpoint	0.1490	0.014754
#15	wespeaker/ecapa512-voxceleb wespeaker_ecapa512	Public checkpoint	0.2235	0.015499
#16	iic/eres2net-en eres_en	Public checkpoint	0.2235	0.014903
#17	speechbrain/ecapa ecapa	Public checkpoint	0.2608	0.021386
#18	wespeaker/ecapa1024-voxceleb wespeaker_ecapa1024	Public checkpoint	0.3055	0.017288
#19	wespeaker/campplus-voxceleb campplus_wespeaker	Public checkpoint	0.4098	0.039531
#20	iic/eres2netv2-zh eresv2_zh	Public checkpoint	1.2034	0.075149
#21	iic/eres2net-large-3dspeaker eres2net_large_3dspeaker	Public checkpoint	1.7139	0.100335
#22	iic/campplus campplus	Public checkpoint	1.7884	0.119188
#23	wespeaker/res34-cnceleb wespeaker_r34	Public checkpoint	3.3905	0.196237
#24	microsoft/unispeech-sat-base-plus-sv unispeech_sat_base_plus	Public checkpoint	7.9993	0.495715
#25	microsoft/wavlm-base-plus-sv wavlm_base	Public checkpoint	8.1967	0.544113