Motion Synthesis On Humanml3D

评估指标

Diversity

FID

Multimodality

R Precision Top3

评测结果

各个模型在此基准测试上的表现结果

					Paper Title	Repository
Language2Pose	7.676	11.02	-	0.486	TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and Texts
Diffuion Motion	23.692	10.21	-	0.735	Diffusion Motion: Generate Text-Guided 3D Human Motion by Diffusion Model	-
Text2Gesture	6.409	5.012	-	0.345	TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and Texts
TM2T	8.589	1.501	2.424	0.729	TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and Texts
T2M	9.175	1.087	2.219	0.736	Generating Diverse and Natural 3D Human Motions From Text	-
TM2D (t2m)	9.513	1.021	4.139	-	TM2D: Bimodality Driven 3D Dance Generation via Music-Text Integration
MAA	8.23	0.774	-	0.676	Make-An-Animation: Large-Scale Text-conditional 3D Human Motion Generation	-
MotionDiffuse	9.410	0.630	1.553	0.782	MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model
MDM	9.559	0.544	2.799	0.611	Human Motion Diffusion Model
MLD	9.724	0.473	2.413	0.772	Executing your Commands via Motion Diffusion in Latent Space
MotionLCM (4-step)	9.607	0.304	2.259	0.798	MotionLCM: Real-time Controllable Motion Generation via Latent Consistency Model
Motion Mamba	9.871	0.281	2.294	0.792	Motion Mamba: Efficient and Long Sequence Motion Generation
Fg-T2M	9.278	0.243	1.614	0.783	Fg-T2M: Fine-Grained Text-Driven Human Motion Generation via Diffusion Model	-
FineMoGen	9.263	0.151	2.696	0.784	FineMoGen: Fine-Grained Spatio-Temporal Motion Generation and Editing
T2M-GPT (τ ∈ U[0, 1])	9.722	0.141	1.831	0.775	T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representations
T2M-GPT (τ = 0)	9.844	0.140	3.285	0.685	T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representations
GraphMotion	9.692	0.116	2.766	0.785	Act As You Wish: Fine-Grained Control of Motion Diffusion Model with Hierarchical Semantic Graphs	-
T2M-GPT (τ = 0.5)	9.761	0.116	1.856	0.775	T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representations
AttT2M	9.700	0.112	2.452	0.786	AttT2M: Text-Driven Human Motion Generation with Multi-Perspective Attention Mechanism
EMDM	9.551	0.112	1.641	0.786	EMDM: Efficient Motion Diffusion Model for Fast and High-Quality Motion Generation

0 of 37 row(s) selected.