Browse State-of-the-Art › Skeleton Based Action Recognition

Skeleton Based Action Recognition

219 papers with code · 34 benchmarks · 30 datasets archive 2025-07-28

Computer VisionNatural Language ProcessingTime Series

Skeleton-based Action Recognition is a computer vision task that involves recognizing human actions from a sequence of 3D skeletal joint data captured from sensors such as Microsoft Kinect, Intel RealSense, and wearable devices. The goal of skeleton-based action recognition is to develop algorithms that can understand and classify human actions from skeleton data, which can be used in various applications such as human-computer interaction, sports analysis, and surveillance.

( Image credit: View Adaptive Neural Networks for High Performance Skeleton-based Human Action Recognition )

Description from the archive archive 2025-07-28.

Benchmarks archive 2025-07-28

34 leaderboard tables shown for this task, 34 with rows (a “benchmark” on this site is a table with at least one row, as on /sota), ordered by row count. “Best model” is the first row in the archive's own order at snapshot; nothing is re-ranked here and metric direction is not recorded in the archive. PwC's Trend sparklines are not in the archive, so that column is omitted. 10 shown of 34 until expanded.

Syntology 3 newer papers were judged to report on this task's benchmarks so far (Claude Sonnet 4.5, reading their tables: a model's judgement, not a result; 1 placed): NTU RGB+D (1 judged, 0 placed), Kinetics-Skeleton dataset (2 judged, 1 placed), MSR Action3D (1 judged, 0 placed). Drawn from 8,886 of 18,637 newer papers checked so far (733 more had nothing to place; 41 could not be read).

DatasetBest model (first row in archive order)PaperCodeSyntologyCompare
NTU RGB+D (135 rows) Hulk(Finetune, ViT-L) Hulk: A Universal Knowledge Translator for Human-Centric Tasks code Syntology official (archive's flag): 13 ran · 13 ran (of which 0 constructed an object rather than computing a result; 4 with no instrument failure: 1 honoured, 0 violated, 3 with no contract checked; 9 where Syntology's instrument failed) · 11 unverified Compare
NTU RGB+D 120 (83 rows) ProtoGCN Revealing Key Details to See Differences: A Novel Prototypical... code — Compare
Kinetics-Skeleton dataset (42 rows) Structured Keypoint Pooling (PPNv2 skeletons+objects) Unified Keypoint-based Action Recognition Framework via Structured... — — Compare
N-UCLA (25 rows) DSCNet (RGB + Pose) A Dense-Sparse Complementary Network for Human Action Recognition... code — Compare
J-HMDB (13 rows) Potion PoTion: Pose MoTion Representation for Action Recognition — — Compare
SBU / SBU-Refine (9 rows) Joint Line Distance On Geometric Features for Skeleton-Based Action Recognition using... code — Compare
SYSU 3D (9 rows) SGN Semantics-Guided Neural Networks for Efficient Skeleton-Based... code Syntology official (archive's flag): 3 ran · 11 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 3 where Syntology's instrument failed) · 2 unverified Compare
UAV-Human (9 rows) HDBN HDBN: A Novel Hybrid Dual-branch Network for Robust Skeleton-based... code — Compare
CAD-120 (8 rows) NGM (5-shot) Neural Graph Matching Networks for Fewshot 3D Action Recognition — — Compare
Florence 3D (7 rows) Deep STGC_K Spatio-Temporal Graph Convolution for Skeleton Based Action Recognition — — Compare
SHREC 2017 track on 3D Hand Gesture Recognition (7 rows) RE-TCN Action Recognition in Real-World Ambient Assisted Living Environment code — Compare
UT-Kinect (7 rows) Temporal Subspace Clustering Subspace Clustering for Action Recognition with Covariance... code — Compare
Varying-view RGB-D Action-Skeleton (7 rows) VS-CNN A Large-scale Varying-view RGB-D Action Dataset for Arbitrary-view... — — Compare
JHMDB (2D poses only) (6 rows) HT-ConvNet Hierarchical Temporal Convolution Network:Towards Privacy-Centric... code — Compare
First-Person Hand Action Benchmark (4 rows) TCN-Summ Domain and View-point Agnostic Hand Action Recognition code — Compare
Gaming 3D (G3D) (4 rows) CNN Action Recognition Based on Joint Trajectory Maps with... — — Compare
H2O (2 Hands and Objects) (4 rows) CHASE(STSA-Net) CHASE: Learning Convex Hull Adaptive Shift for Skeleton-based... code Syntology official (archive's flag): 2 ran · 2 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified Compare
MSR Action3D (4 rows) Distribution of Action Movements + SVM Distribution of Action Movements (DAM): A Descriptor for Human... — — Compare
PKU-MMD (4 rows) RF-Action Making the Invisible Visible: Action Recognition Through Walls and... — — Compare
JHMDB Pose Tracking (3 rows) mgPFF+ft 1st Multigrid Predictive Filter Flow for Unsupervised Learning on Videos code Syntology 4 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 4 where Syntology's instrument failed) · 1 unverified Compare
NTU60-X (3 rows) 4s-ShiftGCN NTU-X: An Enhanced Large-scale Dataset for Improving Pose-based... code — Compare
UPenn Action (3 rows) UNIK UNIK: A Unified Framework for Real-world Skeleton-based Action Recognition code — Compare
UWA3D (3 rows) VA-fusion (aug.) View Adaptive Neural Networks for High Performance Skeleton-based... code Syntology official (archive's flag): 2 ran · 6 ran (of which 1 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 5 where Syntology's instrument failed) · 6 unverified Compare
Drive&Act (2 rows) dyalyt Do You Act Like You Talk? Exploring Pose-based Driver Action... code — Compare
HMDB51 (2 rows) Structured Keypoint Pooling Unified Keypoint-based Action Recognition Framework via Structured... — — Compare
J-HMBD Early Action (2 rows) DR^2N Relational Autoencoder for Feature Extraction code — Compare
MSRC-12 (2 rows) Temporal Subspace Clustering Subspace Clustering for Action Recognition with Covariance... code — Compare
TCG-dataset (2 rows) Bidirectional LSTM Traffic Control Gesture Recognition for Autonomous Vehicles code — Compare
Skeletics-152 (1 row) 4s-ShiftGCN Quo Vadis, Skeleton Action Recognition ? code — Compare
HDM05 (1 row) Temporal Subspace Clustering Subspace Clustering for Action Recognition with Covariance... code — Compare
Kinetics-400 (1 row) STGAT Spatial Temporal Graph Attention Network for Skeleton-Based Action... code — Compare
MSR ActionPairs (1 row) Temporal Subspace Clustering Subspace Clustering for Action Recognition with Covariance... code — Compare
Skeleton-Mimetics (1 row) MS-G3D Quo Vadis, Skeleton Action Recognition ? code — Compare
UCF101 (1 row) Structured Keypoint Pooling Unified Keypoint-based Action Recognition Framework via Structured... — — Compare

Syntology column: samples harvested from the paper's repositories and executed on synthesized fixtures; “ran” is not a correctness claim and does not order the table. Each line reads “N ran (of which C constructed an object rather than computing a result; K with no instrument failure: H honoured, V violated, P with no contract checked; I where Syntology's instrument failed) · U unverified”; the instrument figure counts failures of Syntology's instrument, not of the code. When the archive marks a repository official for the row's paper, the line starts with that repository's state (“official” is the archive's flag, not a verdict on who wrote the code; “community repositories only” when every sample that ran came from a community repository, “official: no sample here; runs from other or unrecorded repositories” when some came from a repository the paper names or has in its text, or from none recorded); hover it for the repositories the samples that ran came from. A dash means no Syntology record for that paper, not a recorded non-run. Each “ran” line links to that paper's sample list. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. For agents: get_harvested_code_for_paper(arxiv_id) lists each paper's samples; how to connect.

Libraries

Not in the archive: the export carries no per-task library table, so there is nothing to show at snapshot 2025-07-28.

Datasets archive 2025-07-28

30 datasets whose archive record lists this task, ordered by the archive's paper count.

Subtasks archive 2025-07-28

No subtask under this task in the archive's task tree.

Parent tasks archive 2025-07-28

Papers archive 2025-07-28

archive papers tagged: 419 · with a code link: 219 · where Syntology ran a sample: 52 (42 with a run with no instrument failure, 10 where every run was a failure of Syntology's instrument) Syntology

Show: all tagged papersonly where code ran (52 of 419 tagged: 42 with a run with no instrument failure, 10 where every run was a failure of Syntology's instrument)

Page 1 of 5: papers 1 to 100 of 419, in archive order: by repositories listed in the archive (most first), then newest first, not by stars (the archive holds no stars, so PwC's “Social” and “Latest” sorts cannot be reproduced). Papers that list no repository come after every paper that lists one.

Papers without a page here are shown as plain text. A Syntology line reads “N ran (of which C constructed an object rather than computing a result; K with no instrument failure: H honoured, V violated, P with no contract checked; I where Syntology's instrument failed) · U unverified”; the figure “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. When the archive marks a repository official for the paper, the line starts with that repository's state (the archive's flag, not a verdict on who wrote the code); hover it for the repositories the samples that ran came from. The first sentence of the abstract is shown for the first 30 papers; every abstract is on its paper's page.

Syntology lines on 30 of the papers shown; no Syntology record for the others (a paper without an arXiv id cannot be joined to the graph, and absence from the graph layer is not a recorded non-run). “Ran” means the sample executed on a synthesized fixture, not that the paper's result was reproduced; each line links to that paper's sample list. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. For agents: get_harvested_code_for_paper(arxiv_id) lists each paper's samples; how to connect.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections