Home News Latest Papers Tutorials Datasets Wiki SOTA LLM Models GPU Leaderboard Events

English

Multimodal Sentiment Analysis On Cmu Mosi

Metrics

Acc-2

Acc-7

Corr

F1

MAE

Results

Performance results of various models on this benchmark

Model Name	Acc-2	Acc-7	Corr	F1	MAE	Paper Title	Repository
TEASEL	87.5	47.52	0.836	85	0.64	TEASEL: A Transformer-Based Speech-Prefixed Language Model
self-M	82.54	45.79	0.795	82.68	0.712	Improving Multimodal Fusion with Hierarchical Mutual Information Maximization for Multimodal Sentiment Analysis
SPECTRA	87.5	-	-	-	-	Speech-Text Dialog Pre-training for Spoken Dialog Understanding with Explicit Cross-Modal Alignment
UniVL + MELTR	85.3	-	0.789	85.4	0.759	MELTR: Meta Loss Transformer for Learning to Fine-tune Video Foundation Models
Tri-TransModality	-	-	-	-	-	TransModality: An End2End Fusion Method with Transformer for Multimodal Sentiment Analysis	-
MMML	90.35	52.72	0.8824	90.35	0.5573	Multimodal Multi-loss Fusion Network for Sentiment Analysis
MMIM	84.14	46.65	0.8	84	0.7	Improving Multimodal Fusion with Hierarchical Mutual Information Maximization for Multimodal Sentiment Analysis
MAG-BERT*	82.37	43.62	0.781	82.5	0.727	Improving Multimodal Fusion with Hierarchical Mutual Information Maximization for Multimodal Sentiment Analysis
UniMSE	86.9	48.68	0.809	86.42	0.691	UniMSE: Towards Unified Multimodal Sentiment Analysis and Emotion Recognition
ALMT	-	49.42	0.805	-	0.683	Learning Language-guided Adaptive Hyper-modality Representation for Multimodal Sentiment Analysis
MCEN	87.35	50.58	0.813	87.48	0.678	10,000+ Times Accelerated Robust Subset Selection (ARSS)	-
VAE-AMDT	84.3	-	-	84.2	0.716	Adversarial Multimodal Domain Transfer for Video-Level Sentiment Analysis	-

0 of 12 row(s) selected.