Motion Synthesis on HumanML3D

Metric: FID (lower is better)

LeaderboardDataset

Loading chart...

Results

Hide extra data

Sort:

#	Model↕	FID▲	Extra Data	Paper	Date↕	Code
1	Motion Anything	0.028	No	Motion Anything: Any to Motion Generation	2025-03-10	Code
2	MotionPCM	0.03	No	MotionPCM: Real-Time Motion Synthesis with Phase...	2025-01-31	-
3	BiPO	0.03	No	BiPO: Bidirectional Partial Occlusion Network fo...	2024-11-28	-
4	DisCoRD (+MoMask)	0.032	No	DisCoRD: Discrete Tokens to Continuous Motion vi...	2024-11-29	-
5	MoMask	0.045	No	MoMask: Generative Masked Modeling of 3D Human M...	2023-11-29	Code
6	BAD (CBS)	0.049	No	BAD: Bidirectional Auto-regressive Diffusion for...	2024-09-17	Code
7	Free-T2M (StableMoFusion)	0.051	No	Free-T2M: Frequency Enhanced Text-to-Motion Diff...	2025-01-30	Code
8	MCM	0.053	No	MCM: Multi-condition Motion Synthesis Framework	2024-04-19	Code
9	BAMM	0.055	No	BAMM: Bidirectional Autoregressive Motion Model	2024-03-28	Code
10	BAD (OAAS)	0.065	No	BAD: Bidirectional Auto-regressive Diffusion for...	2024-09-17	Code
11	DiverseMotion (s=1)	0.07	Yes	DiverseMotion: Towards Diverse Human Motion Gene...	2023-09-04	-
12	DiverseMotion (s=2)	0.072	Yes	DiverseMotion: Towards Diverse Human Motion Gene...	2023-09-04	-
13	MMM (predict length)	0.08	No	MMM: Generative Masked Motion Model	2023-12-06	Code
14	MMM (gt length)	0.089	No	MMM: Generative Masked Motion Model	2023-12-06	Code
15	ReMoDiffuse	0.103	No	ReMoDiffuse: Retrieval-Augmented Motion Diffusio...	2023-04-03	Code
16	ParCo	0.109	No	ParCo: Part-Coordinating Text-to-Motion Synthesis	2024-03-27	Code
17	GUESS	0.109	No	GUESS:GradUally Enriching SyntheSis for Text-Dri...	2024-01-04	Code
18	AttT2M	0.112	No	AttT2M: Text-Driven Human Motion Generation with...	2023-09-02	Code
19	EMDM	0.112	No	EMDM: Efficient Motion Diffusion Model for Fast ...	2023-12-04	Code
20	GraphMotion	0.116	No	-	-	Code
21	T2M-GPT (τ = 0.5)	0.116	No	T2M-GPT: Generating Human Motion from Textual De...	2023-01-15	Code
22	T2M-GPT (τ = 0)	0.14	No	T2M-GPT: Generating Human Motion from Textual De...	2023-01-15	Code
23	T2M-GPT (τ ∈ U[0, 1])	0.141	No	T2M-GPT: Generating Human Motion from Textual De...	2023-01-15	Code
24	FineMoGen	0.151	No	FineMoGen: Fine-Grained Spatio-Temporal Motion G...	2023-12-22	Code
25	Fg-T2M	0.243	No	Fg-T2M: Fine-Grained Text-Driven Human Motion Ge...	2023-09-12	-
26	Motion Mamba	0.281	No	Motion Mamba: Efficient and Long Sequence Motion...	2024-03-12	Code
27	MotionLCM (4-step)	0.304	No	MotionLCM: Real-time Controllable Motion Generat...	2024-04-30	Code
28	MLD	0.473	No	Executing your Commands via Motion Diffusion in ...	2022-12-08	Code
29	MDM	0.544	No	Human Motion Diffusion Model	2022-09-29	Code
30	MotionDiffuse	0.63	No	MotionDiffuse: Text-Driven Human Motion Generati...	2022-08-31	Code
31	MAA	0.774	No	Make-An-Animation: Large-Scale Text-conditional ...	2023-05-16	-
32	TM2D (t2m)	1.021	No	TM2D: Bimodality Driven 3D Dance Generation via ...	2023-04-05	Code
33	T2M	1.087	No	-	-	Code
34	TM2T	1.501	No	TM2T: Stochastic and Tokenized Modeling for the ...	2022-07-04	Code
35	Text2Gesture	5.012	No	TM2T: Stochastic and Tokenized Modeling for the ...	2022-07-04	Code
36	Diffuion Motion	10.21	No	Diffusion Motion: Generate Text-Guided 3D Human ...	2022-10-22	-
37	Language2Pose	11.02	No	TM2T: Stochastic and Tokenized Modeling for the ...	2022-07-04	Code

#1Motion AnythingSOTA
0.028
FID· 2025-03-10
Motion Anything: Any to Motion Generation Code
#2MotionPCM
0.03
FID· 2025-01-31
MotionPCM: Real-Time Motion Synthesis with Phased Consistency Model
#3BiPOSOTA
0.03
FID· 2024-11-28
BiPO: Bidirectional Partial Occlusion Network for Text-to-Motion Synthesis
#4DisCoRD (+MoMask)
0.032
FID· 2024-11-29
DisCoRD: Discrete Tokens to Continuous Motion via Rectified Flow Decoding
#5MoMaskSOTA
0.045
FID· 2023-11-29
MoMask: Generative Masked Modeling of 3D Human Motions Code
#6BAD (CBS)
0.049
FID· 2024-09-17
BAD: Bidirectional Auto-regressive Diffusion for Text-to-Motion Generation Code
#7Free-T2M (StableMoFusion)
0.051
FID· 2025-01-30
Free-T2M: Frequency Enhanced Text-to-Motion Diffusion Model With Consistency Loss Code
#8MCM
0.053
FID· 2024-04-19
MCM: Multi-condition Motion Synthesis Framework Code
#9BAMM
0.055
FID· 2024-03-28
BAMM: Bidirectional Autoregressive Motion Model Code
#10BAD (OAAS)
0.065
FID· 2024-09-17
BAD: Bidirectional Auto-regressive Diffusion for Text-to-Motion Generation Code
#11DiverseMotion (s=1)SOTA
0.07
FID· Extra Data· 2023-09-04
DiverseMotion: Towards Diverse Human Motion Generation via Discrete Diffusion
#12DiverseMotion (s=2)SOTA
0.072
FID· Extra Data· 2023-09-04
DiverseMotion: Towards Diverse Human Motion Generation via Discrete Diffusion
#13MMM (predict length)
0.08
FID· 2023-12-06
MMM: Generative Masked Motion Model Code
#14MMM (gt length)
0.089
FID· 2023-12-06
MMM: Generative Masked Motion Model Code
#15ReMoDiffuseSOTA
0.103
FID· 2023-04-03
ReMoDiffuse: Retrieval-Augmented Motion Diffusion Model Code
#16ParCo
0.109
FID· 2024-03-27
ParCo: Part-Coordinating Text-to-Motion Synthesis Code
#17GUESS
0.109
FID· 2024-01-04
GUESS:GradUally Enriching SyntheSis for Text-Driven Human Motion Generation Code
#18AttT2M
0.112
FID· 2023-09-02
AttT2M: Text-Driven Human Motion Generation with Multi-Perspective Attention Mechanism Code
#19EMDM
0.112
FID· 2023-12-04
EMDM: Efficient Motion Diffusion Model for Fast and High-Quality Motion Generation Code
#20GraphMotion
0.116
FID
No paperCode
#21T2M-GPT (τ = 0.5)SOTA
0.116
FID· 2023-01-15
T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representations Code
#22T2M-GPT (τ = 0)SOTA
0.14
FID· 2023-01-15
T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representations Code
#23T2M-GPT (τ ∈ U[0, 1])SOTA
0.141
FID· 2023-01-15
T2M-GPT: Generating Human Motion from Textual Descriptions with Discrete Representations Code
#24FineMoGen
0.151
FID· 2023-12-22
FineMoGen: Fine-Grained Spatio-Temporal Motion Generation and Editing Code
#25Fg-T2M
0.243
FID· 2023-09-12
Fg-T2M: Fine-Grained Text-Driven Human Motion Generation via Diffusion Model
#26Motion Mamba
0.281
FID· 2024-03-12
Motion Mamba: Efficient and Long Sequence Motion Generation Code
#27MotionLCM (4-step)
0.304
FID· 2024-04-30
MotionLCM: Real-time Controllable Motion Generation via Latent Consistency Model Code
#28MLDSOTA
0.473
FID· 2022-12-08
Executing your Commands via Motion Diffusion in Latent Space Code
#29MDMSOTA
0.544
FID· 2022-09-29
Human Motion Diffusion Model Code
#30MotionDiffuseSOTA
0.63
FID· 2022-08-31
MotionDiffuse: Text-Driven Human Motion Generation with Diffusion Model Code
#31MAA
0.774
FID· 2023-05-16
Make-An-Animation: Large-Scale Text-conditional 3D Human Motion Generation
#32TM2D (t2m)
1.021
FID· 2023-04-05
TM2D: Bimodality Driven 3D Dance Generation via Music-Text Integration Code
#33T2M
1.087
FID
No paperCode
#34TM2TSOTA
1.501
FID· 2022-07-04
TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and Texts Code
#35Text2GestureSOTA
5.012
FID· 2022-07-04
TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and Texts Code
#36Diffuion Motion
10.21
FID· 2022-10-22
Diffusion Motion: Generate Text-Guided 3D Human Motion by Diffusion Model
#37Language2PoseSOTA
11.02
FID· 2022-07-04
TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and Texts Code