Tasks SotA Datasets Papers Methods Submit About

Papers With Code 2

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

Explore

Notable Benchmarks All SotA Datasets Papers Methods

Community

Submit Results About

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

Models/SA-DVAE

SA-DVAE

Reported on 98 benchmarks across 7 tasks · 1 paper · 21 SOTA

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

Computer Vision42 results

VideoonNTU RGB+D
Accuracy (5 unseen classes)· 2024-07-18
82.37
best: 86.49 (TDSM)
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D 120
Harmonic Mean (10 unseen classes)· 2024-07-18
60.42
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonPKU-MMD
Random Split Harmonic Mean· 2024-07-18
54.72
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D
Accuracy (5 unseen classes)· 2024-07-18
82.37
best: 86.49 (TDSM)
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D 120
Harmonic Mean (10 unseen classes)· 2024-07-18
60.42
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonPKU-MMD
Random Split Harmonic Mean· 2024-07-18
54.72
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D
Accuracy (5 unseen classes)· 2024-07-18
82.37
best: 86.49 (TDSM)
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D 120
Harmonic Mean (10 unseen classes)· 2024-07-18
60.42
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonPKU-MMD
Random Split Harmonic Mean· 2024-07-18
54.72
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D 120
Accuracy (10 unseen classes)· 2024-07-18
68.77
best: 79.99 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D 120
Accuracy (24 unseen classes)· 2024-07-18
46.12
best: 65.06 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D 120
Random Split Accuracy· 2024-07-18
50.67
best: 71.38 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonPKU-MMD
Random Split Accuracy· 2024-07-18
66.54
best: 87.8 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D
Accuracy (12 unseen classes)· 2024-07-18
41.38
best: 56.03 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D
Random Split Accuracy· 2024-07-18
84.2
best: 93.17 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D
Harmonic Mean (12 unseen classes)· 2024-07-18
42.56
best: 49.7 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D
Harmonic Mean (5 unseen classes)· 2024-07-18
66.27
best: 68.83 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D
Random Split Harmonic Mean· 2024-07-18
75.27
best: 75.51 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D 120
Harmonic Mean (24 unseen classes)· 2024-07-18
44.5
best: 52.4 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
VideoonNTU RGB+D 120
Random Split Harmonic Mean· 2024-07-18
47.54
best: 50.72 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D 120
Accuracy (10 unseen classes)· 2024-07-18
68.77
best: 79.99 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D 120
Accuracy (24 unseen classes)· 2024-07-18
46.12
best: 65.06 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D 120
Random Split Accuracy· 2024-07-18
50.67
best: 71.38 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonPKU-MMD
Random Split Accuracy· 2024-07-18
66.54
best: 87.8 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D
Accuracy (12 unseen classes)· 2024-07-18
41.38
best: 56.03 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D
Random Split Accuracy· 2024-07-18
84.2
best: 93.17 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D
Harmonic Mean (12 unseen classes)· 2024-07-18
42.56
best: 49.7 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D
Harmonic Mean (5 unseen classes)· 2024-07-18
66.27
best: 68.83 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D
Random Split Harmonic Mean· 2024-07-18
75.27
best: 75.51 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D 120
Harmonic Mean (24 unseen classes)· 2024-07-18
44.5
best: 52.4 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Temporal Action LocalizationonNTU RGB+D 120
Random Split Harmonic Mean· 2024-07-18
47.54
best: 50.72 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D 120
Accuracy (10 unseen classes)· 2024-07-18
68.77
best: 79.99 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D 120
Accuracy (24 unseen classes)· 2024-07-18
46.12
best: 65.06 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D 120
Random Split Accuracy· 2024-07-18
50.67
best: 71.38 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonPKU-MMD
Random Split Accuracy· 2024-07-18
66.54
best: 87.8 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D
Accuracy (12 unseen classes)· 2024-07-18
41.38
best: 56.03 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D
Random Split Accuracy· 2024-07-18
84.2
best: 93.17 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D
Harmonic Mean (12 unseen classes)· 2024-07-18
42.56
best: 49.7 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D
Harmonic Mean (5 unseen classes)· 2024-07-18
66.27
best: 68.83 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D
Random Split Harmonic Mean· 2024-07-18
75.27
best: 75.51 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D 120
Harmonic Mean (24 unseen classes)· 2024-07-18
44.5
best: 52.4 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action LocalizationonNTU RGB+D 120
Random Split Harmonic Mean· 2024-07-18
47.54
best: 50.72 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460

Methodology14 results

Zero-Shot LearningonNTU RGB+D
Accuracy (5 unseen classes)· 2024-07-18
82.37
best: 86.49 (TDSM)
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D 120
Harmonic Mean (10 unseen classes)· 2024-07-18
60.42
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonPKU-MMD
Random Split Harmonic Mean· 2024-07-18
54.72
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D 120
Accuracy (10 unseen classes)· 2024-07-18
68.77
best: 79.99 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D 120
Accuracy (24 unseen classes)· 2024-07-18
46.12
best: 65.06 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D 120
Random Split Accuracy· 2024-07-18
50.67
best: 71.38 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonPKU-MMD
Random Split Accuracy· 2024-07-18
66.54
best: 87.8 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D
Accuracy (12 unseen classes)· 2024-07-18
41.38
best: 56.03 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D
Random Split Accuracy· 2024-07-18
84.2
best: 93.17 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D
Harmonic Mean (12 unseen classes)· 2024-07-18
42.56
best: 49.7 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D
Harmonic Mean (5 unseen classes)· 2024-07-18
66.27
best: 68.83 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D
Random Split Harmonic Mean· 2024-07-18
75.27
best: 75.51 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D 120
Harmonic Mean (24 unseen classes)· 2024-07-18
44.5
best: 52.4 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Zero-Shot LearningonNTU RGB+D 120
Random Split Harmonic Mean· 2024-07-18
47.54
best: 50.72 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460

Robots14 results

Activity RecognitiononNTU RGB+D
Accuracy (5 unseen classes)· 2024-07-18
82.37
best: 86.49 (TDSM)
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D 120
Harmonic Mean (10 unseen classes)· 2024-07-18
60.42
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononPKU-MMD
Random Split Harmonic Mean· 2024-07-18
54.72
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D 120
Accuracy (10 unseen classes)· 2024-07-18
68.77
best: 79.99 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D 120
Accuracy (24 unseen classes)· 2024-07-18
46.12
best: 65.06 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D 120
Random Split Accuracy· 2024-07-18
50.67
best: 71.38 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononPKU-MMD
Random Split Accuracy· 2024-07-18
66.54
best: 87.8 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D
Accuracy (12 unseen classes)· 2024-07-18
41.38
best: 56.03 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D
Random Split Accuracy· 2024-07-18
84.2
best: 93.17 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D
Harmonic Mean (12 unseen classes)· 2024-07-18
42.56
best: 49.7 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D
Harmonic Mean (5 unseen classes)· 2024-07-18
66.27
best: 68.83 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D
Random Split Harmonic Mean· 2024-07-18
75.27
best: 75.51 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D 120
Harmonic Mean (24 unseen classes)· 2024-07-18
44.5
best: 52.4 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Activity RecognitiononNTU RGB+D 120
Random Split Harmonic Mean· 2024-07-18
47.54
best: 50.72 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460

Natural Language Processing14 results

3D Action RecognitiononNTU RGB+D
Accuracy (5 unseen classes)· 2024-07-18
82.37
best: 86.49 (TDSM)
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D 120
Harmonic Mean (10 unseen classes)· 2024-07-18
60.42
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononPKU-MMD
Random Split Harmonic Mean· 2024-07-18
54.72
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D 120
Accuracy (10 unseen classes)· 2024-07-18
68.77
best: 79.99 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D 120
Accuracy (24 unseen classes)· 2024-07-18
46.12
best: 65.06 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D 120
Random Split Accuracy· 2024-07-18
50.67
best: 71.38 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononPKU-MMD
Random Split Accuracy· 2024-07-18
66.54
best: 87.8 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D
Accuracy (12 unseen classes)· 2024-07-18
41.38
best: 56.03 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D
Random Split Accuracy· 2024-07-18
84.2
best: 93.17 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D
Harmonic Mean (12 unseen classes)· 2024-07-18
42.56
best: 49.7 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D
Harmonic Mean (5 unseen classes)· 2024-07-18
66.27
best: 68.83 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D
Random Split Harmonic Mean· 2024-07-18
75.27
best: 75.51 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D 120
Harmonic Mean (24 unseen classes)· 2024-07-18
44.5
best: 52.4 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
3D Action RecognitiononNTU RGB+D 120
Random Split Harmonic Mean· 2024-07-18
47.54
best: 50.72 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460

Time Series14 results

Action RecognitiononNTU RGB+D
Accuracy (5 unseen classes)· 2024-07-18
82.37
best: 86.49 (TDSM)
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D 120
Harmonic Mean (10 unseen classes)· 2024-07-18
60.42
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononPKU-MMD
Random Split Harmonic Mean· 2024-07-18
54.72
SOTA
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D 120
Accuracy (10 unseen classes)· 2024-07-18
68.77
best: 79.99 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D 120
Accuracy (24 unseen classes)· 2024-07-18
46.12
best: 65.06 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D 120
Random Split Accuracy· 2024-07-18
50.67
best: 71.38 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononPKU-MMD
Random Split Accuracy· 2024-07-18
66.54
best: 87.8 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D
Accuracy (12 unseen classes)· 2024-07-18
41.38
best: 56.03 (TDSM)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D
Random Split Accuracy· 2024-07-18
84.2
best: 93.17 (PGFA)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D
Harmonic Mean (12 unseen classes)· 2024-07-18
42.56
best: 49.7 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D
Harmonic Mean (5 unseen classes)· 2024-07-18
66.27
best: 68.83 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D
Random Split Harmonic Mean· 2024-07-18
75.27
best: 75.51 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D 120
Harmonic Mean (24 unseen classes)· 2024-07-18
44.5
best: 52.4 (MSF-GZSSAR)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460
Action RecognitiononNTU RGB+D 120
Random Split Harmonic Mean· 2024-07-18
47.54
best: 50.72 (SA-DVAE + augmented text)
SA-DVAE: Improving Zero-Shot Skeleton-Based Action Recognition by Disentangled Variational Autoencoders arXiv:2407.13460