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

Multimodal Emotion Recognition research benchmarks – Page 2 · SOTA2 Research

Multimodal Emotion Recognition

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Segment Level Attention + Bi-Modal Transformer

76.81Accuracy

5

Jun 9, 2026

82.06Acc (2 Class)

5

Feb 26, 2026

Weighted blending of supervision signals

70.05F1-0.25MSE

4

Jun 9, 2026

Memory Bear AI Business Dataset

68.4Accuracy

4

Mar 25, 2026

MER-MULTI (val)

weighted blending supervision signals

70.05F1-0.25MSE

3

Jun 9, 2026

90.36F1 Score

1

Jun 9, 2026

Segment Level Attention + Bi-Modal Transformer

0.743Accuracy

1

Jun 9, 2026

Hybrid latent space fusion

93Accuracy

1

Jun 9, 2026

Page 2 of 2

Popular datasets

IEMOCAP MER CREMA-D CH-SIMS CMU-MOSEI

Follow for updates

Join SOTA2 Contact