Benchmark Scenario
Overlap Robustness
Overlapping speakers. The scenario score aggregates its trial sets using the fixed OpenSVBench reporting protocol.
18 ranked systems
2 trial sets
10.1383 leading EER
Scenario Ranking
Systems on Overlap Robustness
Evidence cells list trial-set rank and EER in this order: AliMeeting Overlap · CHiME-6 Overlap.
| Rank | System | Scenario EER ↓ | Scenario minDCF ↓ | Trial-Set Evidence Rank / EER |
|---|---|---|---|---|
| #1 |
wespeaker/w2vbert2
vb2-w2vbert
|
10.1383 | 0.467346 | |
| #2 |
palabraai/redimnet2-b6
redimnet2
|
10.6981 | 0.524446 | |
| #3 |
wespeaker/samresnet100
vb2-sam100
|
11.0912 | 0.502019 | |
| #4 |
wespeaker/res293-voxceleb
wespeaker_resnet293
|
12.3294 | 0.554516 | |
| #5 |
wespeaker/res152-voxceleb
wespeaker_resnet152
|
12.9553 | 0.565796 | |
| #6 |
iic/eres2net-en
eres_en
|
13.7639 | 0.580907 | |
| #7 |
wespeaker/res34-voxceleb
wespeaker_resnet34_voxceleb
|
13.9414 | 0.591408 | |
| #8 |
wespeaker/campplus-voxceleb
campplus_wespeaker
|
14.0816 | 0.622128 | |
| #9 |
iic/eres2net-large-3dspeaker
eres2net_large_3dspeaker
|
14.3667 | 0.552132 | |
| #10 |
wespeaker/ecapa1024-voxceleb
wespeaker_ecapa1024
|
14.6411 | 0.618143 | |
| #11 |
speechbrain/ecapa
ecapa
|
14.8522 | 0.616222 | |
| #12 |
iic/eres2netv2-zh
eresv2_zh
|
14.8800 | 0.543251 | |
| #13 |
wespeaker/ecapa512-voxceleb
wespeaker_ecapa512
|
14.9489 | 0.650355 | |
| #14 |
wespeaker/res34-cnceleb
wespeaker_r34
|
16.4122 | 0.671276 | |
| #15 |
iic/campplus
campplus
|
18.6371 | 0.617367 | |
| #16 |
speechbrain/xvector
xvector
|
23.2363 | 0.916343 | |
| #17 |
microsoft/wavlm-base-plus-sv
wavlm_base
|
26.0906 | 0.926016 | |
| #18 |
microsoft/unispeech-sat-base-plus-sv
unispeech_sat_base_plus
|
26.5022 | 0.944448 |