Our new X account is live! Follow @wizwand_team for updates

Multimodal Emotion Recognition on MMEVerse-Bench (test)

0.6212CAER

HumanOmni-7B

Updated 4d ago

Evaluation Results

Method	Links
HumanOmni-7B 2026.01		0.6212	0.4133	0.7433	0.7542	0.4407	0.0845	0.3877	0.3042	0.5254	0.45
Emotion-LLaMAv2 2026.01		0.5821	0.6335	0.707	0.5073	0.3588	0.091	0.7471	0.656	0.6436	0.6663
Emotion-LLaMAv2 2026.01		0.581	0.5861	0.7087	0.4965	0.3406	0.0921	0.7322	0.6343	0.615	0.6605
Qwen2.5-Omni-7B 2026.01		0.408	0.4502	0.5023	0.323	0.077	0.077	0.6567	0.4829	0.4691	0.5273
Emotion-LLaMA 2026.01		0.4054	0.3542	0.4686	0.3067	0.2611	0.0767	0.5954	0.3809	0.4811	0.5059
LLaMA3.2-90B-Vision 2026.01		0.3792	0.3309	0.469	0.3246	0.2156	0.0717	0.6383	0.3786	0.4349	0.5066
Video-LLaMA 2026.01		0.3762	0.3247	0.3759	0.2436	0.2253	0.0758	0.5184	0.2262	0.4483	0.3526
HumanOmniV2-9B 2026.01		0.3652	0.4213	0.4413	0.2001	0.2867	0.0836	0.4222	0.338	0.367	0.5006
Qwen2.5-VL-72B 2026.01		0.3261	0.3333	0.3195	0.2175	0.1167	0.0732	0.6513	0.5623	0.4885	0.4596
AffectGPT 2026.01		0.3045	0.2761	0.4361	0.2626	0.2318	0.0722	0.5893	0.3795	0.4635	0.5411
Video-LLaVA 2026.01		0.2787	0.3124	0.3729	0.2257	0.2203	0.0707	0.4314	0.2202	0.3061	0.3878
Video-LLaMA2 2026.01		0.2173	0.3044	0.3853	0.2751	0.2537	0.0768	0.5762	0.2562	0.1302	0.3983