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

Multi-speaker Automatic Speech Recognition research benchmarks · SOTA2 Research

Multi-speaker Automatic Speech Recognition

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

AliMeeting (test)

SpeakerLM (7639h)

16.05cpCER

30

Jun 12, 2026

DM-ASR (S2SND) (CN+EN 2900h)

22.07tcpWER

14

Jun 11, 2026

Aishell4 (eval)

SpeakerLM* (7638 hours)

17.17Character Error Rate (CER)

13

Jun 12, 2026

DM-ASR (S2SND) (CN+EN 2900h)

16.4cpWER

12

Apr 27, 2026

24.62CP-WER

11

Mar 12, 2026

AliMeeting (eval)

Temporal Interleave

25.56CER

10

Jun 12, 2026

14.34CP-WER

8

Feb 26, 2026

9.88CP-WER

7

Feb 26, 2026

34.48CP-WER

7

Feb 26, 2026

28.4WER

7

Feb 26, 2026

Popular datasets

AMI AliMeeting Libri2Mix LibriCSS NOTSOFAR

Follow for updates

Join SOTA2 Contact