Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Speaker Diarization research benchmarks – Page 3 · SOTA2 Research

Speaker Diarization

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

10.272DER

5

Feb 26, 2026

10.18DER

5

Feb 26, 2026

8.325DER

5

Feb 26, 2026

AliMeeting (eval)

3.96DER (%)

5

Feb 26, 2026

NIST RT-03 (LDC2007S10) (eval)

0.88Speaker Error Rate

5

Feb 26, 2026

CH109 CHAES two-speaker (eval)

0.023Speaker Error Rate

5

Feb 26, 2026

CHAES LDC97S42 (eval)

2.48Diarization Error Rate (DER)

5

Feb 26, 2026

RT03 LDC2007S10 (test)

3.1DER

5

Feb 26, 2026

CH109 CHAES (test)

5.61DER

5

Feb 26, 2026

CHAES LDC97S42 (val)

5.04DER

5

Feb 26, 2026

DIHARD II (test)

26.1DER

5

Feb 26, 2026

13.83DER

4

Jun 30, 2026

19.8DER

4

Jun 24, 2026

Average (AIS, AliM, AMI, DH3, MSD, NSF, RAMC, VoxC)

MS-SphereVBx-PF

12.48DER (%)

4

Jun 24, 2026

MSDWild many (val)

38.4DER micro

4

Jun 9, 2026

Dual-Strategy-Enhanced ConBiMamba

19.2DER (%)

4

Feb 26, 2026

CALLHOME overlap ratio 11.8%

11.53DER

4

Feb 26, 2026

50-conversation synthetic code-switching benchmark

0.693ARI (mean)

3

May 4, 2026

Speaker Diarization Dataset

Pyannote Diarization V2.1

18.36DER

3

Apr 2, 2026

5.5tcpWER

3

Feb 26, 2026

VoxConverse (Eval)

4.27DER (Overall)

3

Feb 26, 2026

1.77Diarization Error (Mix-Headset)

3

Feb 26, 2026

DebateIndonesianLarge (test)

Indo Adapted (25h)

77.78Precision

3

Feb 26, 2026

4.6DER (%)

3

Feb 26, 2026

Simulated mixtures beta=5, overlap ratio 11.1%

19.69DER

3

Feb 26, 2026

Page 3 of 4

Popular datasets

AMI CALLHOME DIHARD AliMeeting VoxConverse

Follow for updates

Join SOTA2 Contact