AI Paper Atlas

Research map / Benchmarks / NTU RGB+D 60 X-Sub

NTU RGB+D 60 X-Sub leaderboard

10 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on NTU RGB+D 60 X-Sub. Best accuracy: 97.9 (HP-Net, TPAMI 2025).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

NTU RGB+D 60 X-Sub · accuracy

10 papers, each with its best reported number for its own method; higher is better.

#accuracy ↑MethodPaperTopic
197.9HP-NetHeatmap Pooling Network for Action Recognition From RGB Videos
TPAMI 2025 · Table 1
Action recognition
295.2SUGAR
LLaMA2-7B
SUGAR: Learning Skeleton Representation with Visual-Motion Knowledge for Action Recognition
AAAI 2026 · Table 1
Action recognition
393.7OursSkeleton-based Action Recognition with Non-linear Dependency Modeling and Hilbert-Schmidt Independence Criterion
AAAI 2025 · Table: NTU RGB+D 60
Molecules and proteins
493.4HD-GCN 6-way ensemble
CTR-GCN backbone
Hierarchically Decomposed Graph Convolutional Networks for Skeleton-Based Action Recognition
ICCV 2023 · Appendix table
Knowledge and temporal graphs
593.3STC-Net (6-ensemble)Leveraging Spatio-Temporal Dependency for Skeleton-Based Action Recognition
ICCV 2023 · Table 1
Knowledge and temporal graphs
693.1BlockGCNBlockGCN: Redefine Topology Awareness for Skeleton-Based Action Recognition
CVPR 2024 · Table 2
Knowledge and temporal graphs
792.9CTR-GCN w/ Koopman Pooling
CTR-GCN
Neural Koopman Pooling: Control-Inspired Temporal Dynamics Encoding for Skeleton-Based Action Recognition
CVPR 2023 · Table 1
Action recognition
892.1EfficientGCN-B4Constructing Stronger and Faster Baselines for Skeleton-Based Action Recognition
TPAMI 2022 · Table (NTU 60 SOTA)
Knowledge and temporal graphs
991.6Ours (multi-stream)Learning Skeletal Graph Neural Networks for Hard 3D Pose Estimation
ICCV 2021 · Table 7
Human pose estimation
1091.5MS-G3D NetDisentangling and Unifying Graph Convolutions for Skeleton-Based Action Recognition
CVPR 2020 · Table, NTU RGB+D 60
Knowledge and temporal graphs

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.