notesum.ai

Published at November 21

SpikEmo: Enhancing Emotion Recognition With Spiking Temporal Dynamics in Conversations

cs.MM

Released Date: November 21, 2024

Authors: Xiaomin Yu¹, Feiyang Wang², Ziyue Qiao¹

Aff.: ¹Great Bay University, Guangdong, China; ²Central South University, Changsha, Hunan, China

Arxiv: http://arxiv.org/abs/2411.13917v1

Refer to caption

Models	MELD								IEMOCAP
Models	Neutral	Surprise	Fear	Sadness	Joy	Disgust	Angry	W-F1	Happiness	Sadness	Neutral	Anger	Excitement	Frustration	W-F1
BC-LSTM [1]	73.80	47.70	5.40	25.10	51.30	5.20	38.40	55.90	34.43	60.87	51.81	56.73	57.95	58.92	54.95
DialogueRNN [2]	76.23	49.59	0.00	26.33	54.55	0.81	46.76	58.73	33.18	78.80	59.21	65.28	71.86	58.91	62.75
DialogueGCN [3]	76.02	46.37	0.98	24.32	53.62	1.22	43.03	57.52	51.87	76.76	56.76	62.26	72.71	58.04	63.16
IterativeERC [4]	77.52	53.65	3.31	23.62	56.63	19.38	48.88	60.72	53.17	77.19	61.31	61.45	69.23	60.92	64.37
CESTa [5]	-	-	-	-	-	-	-	58.36	47.70	80.82	64.76	63.41	75.95	62.65	67.10
RGAT-ERC [6]	-	-	-	-	-	-	-	60.91	51.62	77.32	65.42	63.01	67.95	61.23	65.22
QMNN [7]	77.00	49.76	0.00	16.50	52.08	0.00	43.17	58.00	39.71	68.30	55.29	62.58	66.71	62.19	59.88
MMGCN [8]	-	-	-	-	-	-	-	58.65	42.34	78.67	61.73	69.00	74.33	62.32	66.22
MVN [9]	76.65	53.18	11.70	21.82	53.62	21.86	42.55	59.03	55.75	73.30	61.88	65.96	69.50	64.21	65.44
MM-DFN [10]	77.76	50.69	-	22.93	54.78	-	47.82	59.46	42.22	78.98	66.42	69.77	75.56	66.33	68.18
MultiEMO [11]	78.22	56.52	21.78	39.14	62.02	23.76	52.91	64.43	52.55	83.44	66.42	66.07	74.02	63.04	68.33
GA2MIF [12]	76.92	49.08	-	27.18	51.87	-	48.52	58.94	46.15	84.50	68.38	70.29	75.99	66.49	70.00
SpikEmo	79.53	58.40	21.95	40.00	63.89	28.85	54.20	65.92	58.16	82.17	72.61	66.67	79.01	64.82	71.50