TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/TM2T

TM2T

Reported on 36 benchmarks across 5 tasks · 1 paper · 36 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Audio16 results

  • 10-shot image generationonHumanML3D
    Diversity· 2022-07-04
    8.589
    best: 23.692 (Diffuion Motion)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 10-shot image generationonHumanML3D
    FID· 2022-07-04
    1.501
    best: 0.028 (Motion Anything)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 10-shot image generationonHumanML3D
    Multimodality· 2022-07-04
    2.424
    best: 4.139 (TM2D (t2m))
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 10-shot image generationonHumanML3D
    R Precision Top3· 2022-07-04
    0.729
    best: 0.842 (MotionPCM)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 10-shot image generationonKIT Motion-Language
    Diversity· 2022-07-04
    9.473
    best: 11.198 (T2M-GPT (τ = 0))
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 10-shot image generationonKIT Motion-Language
    FID· 2022-07-04
    3.599
    best: 0.131 (Motion Anything)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 10-shot image generationonKIT Motion-Language
    Multimodality· 2022-07-04
    3.292
    best: 3.627 (GraphMotion)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 10-shot image generationonKIT Motion-Language
    R Precision Top3· 2022-07-04
    0.587
    best: 0.803 (BiPO)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 3D Human Pose TrackingonHumanML3D
    Diversity· 2022-07-04
    8.589
    best: 23.692 (Diffuion Motion)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 3D Human Pose TrackingonHumanML3D
    FID· 2022-07-04
    1.501
    best: 0.028 (Motion Anything)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 3D Human Pose TrackingonHumanML3D
    Multimodality· 2022-07-04
    2.424
    best: 4.139 (TM2D (t2m))
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 3D Human Pose TrackingonHumanML3D
    R Precision Top3· 2022-07-04
    0.729
    best: 0.842 (MotionPCM)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 3D Human Pose TrackingonKIT Motion-Language
    Diversity· 2022-07-04
    9.473
    best: 11.198 (T2M-GPT (τ = 0))
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 3D Human Pose TrackingonKIT Motion-Language
    FID· 2022-07-04
    3.599
    best: 0.131 (Motion Anything)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 3D Human Pose TrackingonKIT Motion-Language
    Multimodality· 2022-07-04
    3.292
    best: 3.627 (GraphMotion)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • 3D Human Pose TrackingonKIT Motion-Language
    R Precision Top3· 2022-07-04
    0.587
    best: 0.803 (BiPO)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696

Computer Vision12 results

  • Pose TrackingonHumanML3D
    Diversity· 2022-07-04
    8.589
    best: 23.692 (Diffuion Motion)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Pose TrackingonHumanML3D
    FID· 2022-07-04
    1.501
    best: 0.028 (Motion Anything)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Pose TrackingonHumanML3D
    Multimodality· 2022-07-04
    2.424
    best: 4.139 (TM2D (t2m))
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Pose TrackingonHumanML3D
    R Precision Top3· 2022-07-04
    0.729
    best: 0.842 (MotionPCM)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Pose TrackingonKIT Motion-Language
    Diversity· 2022-07-04
    9.473
    best: 11.198 (T2M-GPT (τ = 0))
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Pose TrackingonKIT Motion-Language
    FID· 2022-07-04
    3.599
    best: 0.131 (Motion Anything)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Pose TrackingonKIT Motion-Language
    Multimodality· 2022-07-04
    3.292
    best: 3.627 (GraphMotion)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Pose TrackingonKIT Motion-Language
    R Precision Top3· 2022-07-04
    0.587
    best: 0.803 (BiPO)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion CaptioningonHumanML3D
    BERTScore· 2022-07-04
    37.8
    best: 40.3 (ST-MLP)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion CaptioningonHumanML3D
    BLEU-4· 2022-07-04
    22.3
    best: 25 (ST-MLP)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion CaptioningonKIT Motion-Language
    BERTScore· 2022-07-04
    23
    best: 42.1 (MLP+GRU)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion CaptioningonKIT Motion-Language
    BLEU-4· 2022-07-04
    18.4
    best: 25.4 (MLP+GRU)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696

Computer Code8 results

  • Motion SynthesisonHumanML3D
    Diversity· 2022-07-04
    8.589
    best: 23.692 (Diffuion Motion)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion SynthesisonHumanML3D
    FID· 2022-07-04
    1.501
    best: 0.028 (Motion Anything)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion SynthesisonHumanML3D
    Multimodality· 2022-07-04
    2.424
    best: 4.139 (TM2D (t2m))
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion SynthesisonHumanML3D
    R Precision Top3· 2022-07-04
    0.729
    best: 0.842 (MotionPCM)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion SynthesisonKIT Motion-Language
    Diversity· 2022-07-04
    9.473
    best: 11.198 (T2M-GPT (τ = 0))
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion SynthesisonKIT Motion-Language
    FID· 2022-07-04
    3.599
    best: 0.131 (Motion Anything)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion SynthesisonKIT Motion-Language
    Multimodality· 2022-07-04
    3.292
    best: 3.627 (GraphMotion)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696
  • Motion SynthesisonKIT Motion-Language
    R Precision Top3· 2022-07-04
    0.587
    best: 0.803 (BiPO)
    SOTA
    TM2T: Stochastic and Tokenized Modeling for the Reciprocal Generation of 3D Human Motions and TextsarXiv:2207.01696