Who leads the VoxCeleb1-O benchmark?

ResNet-34 (AM-Softmax, VoxCeleb2) currently leads VoxCeleb1-O with a score of 1.18 on Eer.

What is the state-of-the-art score on VoxCeleb1-O?

The state-of-the-art result on VoxCeleb1-O is 1.18 (Eer), achieved by ResNet-34 (AM-Softmax, VoxCeleb2) as of 2026.

How many models are tracked on VoxCeleb1-O?

Codesota tracks 3 models on VoxCeleb1-O.

When was the VoxCeleb1-O leaderboard last updated?

The VoxCeleb1-O leaderboard on Codesota includes results through 2026.

Codesota · Benchmark · VoxCeleb1-OHome/Leaderboards/Audio & Speech/Speaker Verification/VoxCeleb1-O

Unknown

VoxCeleb1-O.

Name: VoxCeleb1-O Benchmark Results
Creator: Unknown
Published: 2026-01-01
License: https://creativecommons.org/licenses/by/4.0/

The original VoxCeleb1 speaker-verification test trials (~37k pairs from 40 speakers). EER (Equal Error Rate, lower=better) is the standard metric. SOTA systems are well below 1%.

Paper ↗Leaderboard ↓

§ 01 · Leaderboard

Results by metric.

Only 3 models on this benchmark

Help build the community leaderboard — submit your model results.

Found a wrong score or missing run?

Use row edits to send a sourced correction into moderation.

Add / edit result ↗Report issue ↗

Eer

Eer is the reported evaluation metric for VoxCeleb1-O. Codesota tracks published model scores on this metric so readers can compare state-of-the-art results across sources and model families.

Higher is better

Trust tiers for Eerverifiedpapervendorcommunityunverified

Rank	Model	Trust	Score	Year	Links	Fix
01	ResNet-34 (AM-Softmax, VoxCeleb2) ResNet-34 AM-Softmax baseline on Vox1-O. seed — verify.	paper	1.18	2026	Source ↗	Looks wrong?
02	ECAPA-TDNN ECAPA-TDNN trained on VoxCeleb2, Vox1-O EER %. Desplanques et al. Interspeech 2020.	paper	0.87	2026	Source ↗	Looks wrong?
03	WavLM Large (SV) WavLM Large + ECAPA, VoxCeleb1-O, EER %. Chen et al. 2022.	paper	0.38	2026	Source ↗	Looks wrong?

§ 04 · Submit a result

Add to the leaderboard.

← Back to Speaker Verification