基于多头自注意力机制与MLP-Interactor的多模态情感分析

林宜山,左景,卢树华

Multimodal sentiment analysis based on multi-head self-attention mechanism and MLP-Interactor

Yishan LIN,Jing ZUO,Shuhua LU

表 3 在CMU-MOSEI数据集上和其他基准模型性能的对比结果

Tab.3 Comparison of performance on CMU-MOSEI dataset with other benchmark models

模型 MAE corr A2 A7 F1 TFN[40](2017) 0.593 0.700 —/82.5 50.2 —/82.1 LMF[41](2018) 0.623 0.677 —/82.0 48.0 —/82.1 MulT[20](2019) 0.580 0.703 —/82.5 51.8 —/82.3 BBFN[11](2021) 0.529 0.767 —/86.2 54.8 —/86.1 Self-MM[24](2021) 0.530 0.765 82.81/85.17 — 82.53/85.30 MISA[7](2020) 0.555 0.756 83.6/85.5 52.2 83.8/85.3 MAG-BERT[43](2020) 0.543 0.755 82.51/84.82 — 82.77/84.71 CubeMLP[31](2022) 0.529 0.760 —/85.1 54.9 —/84.5 PS-Mixer[30](2023) 0.537 0.765 83.1/86.1 53.0 83.1/86.1 MTSA[44](2022) 0.541 0.774 —/85.5 52.9 —/85.3 AOBERT[10](2023) 0.515 0.763 84.9/86.2 54.5 85.0/85.9 TETFN[25](2023) TMRN[45](2023) 0.551 0.535 0.748 0.762 84.25/85.18 83.39/86.19 — 53.65 84.18/85.27 83.67/86.08 MTAMW[46](2024) 0.525 0.782 83.09/86.49 53.73 83.48/86.45 MIBSA[47](2024) 0.568 0.753 —/86.70 52.40 —/85.80 FRDIN[48](2024) 0.525 0.778 83.30/86.30 54.40 83.70/86.20 CRNet[49](2024) 0.541 0.771 —/86.20 53.80 —/86.10 本文模型 0.512 0.794 83.0/86.8 54.5 82.5/86.8