基于多视角解耦融合的多模态会话情感识别

张婷婷,朱小飞

Multi-view decoupled fusion for multimodal emotion recognition in conversation

Tingting ZHANG,Xiaofei ZHU

表 2 不同模型在IEMOCAP数据集上的性能对比结果

Tab.2 Performance comparison of different models on IEMOCAP dataset

模型 F1% Acc% W-F1% 高兴 悲伤 中性 愤怒 兴奋 沮丧 DialogueRNN# 32.20 80.26 57.89 62.82 73.87 59.76 63.52 62.89 MMGCN# 45.14 77.16 64.36 68.82 74.71 61.40 66.36 66.26 CTNet# 51.30 79.90 65.80 67.20 78.70 58.80 68.00 67.50 MM-DFN$ 42.22 78.98 66.42 69.77 75.56 66.33 68.21 68.18 SCMM$ 45.37 78.76 63.54 66.05 76.70 66.18 — 67.53 CMCF-SRNet$ 52.20 80.90 68.80 70.30 76.70 61.60 70.50 69.60 MultiDAG$ 49.65 81.40 69.53 70.33 71.61 66.94 69.11 69.08 FrameERC$ 56.19 80.88 67.69 63.16 78.58 69.73 70.79 70.67 AdaIGN$ 53.04 81.47 71.26 65.87 76.34 67.79 70.49 70.74 MDF 56.76* 83.98* 74.84* 66.65 77.66 69.61* 72.87* 73.05*