Text-to-motion retrieval on HumanML3D 1.0 (test) benchmark leaderboard · SOTA2 Research

Text-to-motion retrieval on HumanML3D 1.0 (test)

68.58R@1LAVIMO

Evaluation Results

Method	Links
LAVIMO2024.03		68.58	81.04	85.02	88.77	92.58	1.01
TMR2024.03		67.16	81.32	86.81	91.43	95.36	1.04
LAVIMO2024.03		66.74	78.98	83.67	87.29	91.66	1.02
Guo et al.2024.03		52.48	71.05	80.65	89.86	96.58	1.39
LAVIMO2024.03		51	63	69	76	80	1.5
LAVIMO2024.03		50	71	79	86	90	1
MotionCLIP2024.03		46.24	60.25	68.93	80.47	91.35	1.88
TEMOS2024.03		40.49	53.52	61.14	70.96	84.15	2.33
TMR2024.03		34	56	61	68	76	2
MotionCLIP2024.03		21	31	37	49	65	6
TEMOS2024.03		20	33	37	47	62	6
Guo et al.2024.03		13	27	39	51	72	5
LAVIMO2024.03		12.94	17.38	23.63	30.13	42.46	16
TMR2024.03		11.6	15.39	20.5	27.72	38.52	19
LAVIMO2024.03		11.32	15.01	20.91	27.93	40.97	17
MotionCLIP2024.03		7.22	10.58	13.48	19.07	23.65	55
LAVIMO2024.03		6.37	11.84	15.6	21.95	33.67	24
LAVIMO2024.03		5.93	10.89	14.67	20.78	32.47	24
TMR2024.03		5.68	10.59	14.04	20.34	30.94	28
Guo et al.2024.03		5.3	7.83	10.75	14.59	22.51	54
TEMOS2024.03		5.21	8.22	11.14	15.09	22.12	79
MotionCLIP2024.03		2.33	5.85	8.93	12.77	18.14	103
TEMOS2024.03		2.12	4.09	5.87	8.26	13.52	173
Guo et al.2024.03		1.8	3.42	4.79	7.12	12.47	81