mlsanityData to 14 Oct 2026

3D Skeleton-based Few-shot Action Recognition with JEANIE is not so Na"ive

arXiv 2112.12668 ↗First posted 23 Dec 2021cs.CV cs.HC cs.LGcited by 10

Not on the map. Why?

Abstract (arXiv)

In this paper, we propose a Few-shot Learning pipeline for 3D skeleton-based action recognition by Joint tEmporal and cAmera viewpoiNt alIgnmEnt (JEANIE). To factor out misalignment between query and support sequences of 3D body joints, we propose an advanced variant of Dynamic Time Warping which jointly models each smooth path between the query and support frames to achieve simultaneously the best alignment in the temporal and simulated camera viewpoint spaces for end-to-end learning under the limited few-shot training data. Sequences are encoded with a temporal block encoder based on Simple Spectral Graph Convolution, a lightweight linear Graph Neural Network backbone (we also include a setting with a transformer). Finally, we propose a similarity-based loss which encourages the alignment of sequences of the same class while preventing the alignment of unrelated sequences. We demonstrate state-of-the-art results on NTU-60, NTU-120, Kinetics-skeleton and UWA3D Multiview Activity II.

Key facts. Cited by 10 papers in the map (data through 14 Oct 2026). Builds on 45 of its 109 references that are in the map. First posted on arXiv 23 Dec 2021.

Builds on

45 of 109 references.

References · 45

  1. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  2. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,361
  3. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,897
  4. 2017Quo Vadis, Action Recognition? A New Model and the Kinetics Datasetcited by 3,270
  5. 2017Prototypical Networks for Few-shot Learningcited by 3,015
  6. 2016Matching Networks for One Shot Learningcited by 2,255
  7. 2017The Kinetics Human Action Video Datasetcited by 1,909
  8. 2019Simplifying Graph Convolutional Networkscited by 1,477
  9. 2018Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognitioncited by 1,209
  10. 2017Learning to Compare: Relation Network for Few-Shot Learningcited by 1,143
  11. 2016Realtime Multi-Person 2D Pose Estimation using Part Affinity Fieldscited by 1,064
  12. 2018Predict then Propagate: Graph Neural Networks meet Personalized PageRankcited by 993
33 more
  1. 2016NTU RGB+D: A Large Scale Dataset for 3D Human Activity Analysiscited by 686
  2. 2019NTU RGB+D 120: A Large-Scale Benchmark for 3D Human Activity Understandingcited by 533
  3. 2014Long-term Recurrent Convolutional Networks for Visual Recognition and Descriptioncited by 484
  4. 2017A simple yet effective baseline for 3d human pose estimationcited by 401
  5. 2016Convolutional Two-Stream Network Fusion for Video Action Recognitioncited by 384
  6. 2020Disentangling and Unifying Graph Convolutions for Skeleton-Based Action Recognitioncited by 343
  7. 2017Temporal Segment Networks for Action Recognition in Videoscited by 279
  8. 2017Soft-DTW: a Differentiable Loss Function for Time-Seriescited by 271
  9. 2019Actional-Structural Graph Convolutional Networks for Skeleton-based Action Recognitioncited by 228
  10. 2019An Attention Enhanced Graph Convolutional LSTM Network for Skeleton-Based Action Recognitioncited by 151
  11. 2017A New Representation of Skeleton Sequences for 3D Action Recognitioncited by 126
  12. 2019Few-Shot Video Classification via Temporal Alignmentcited by 125
  13. 2020Adversarial Feature Hallucination Networks for Few-Shot Learningcited by 110
  14. 2017View Adaptive Recurrent Neural Networks for High Performance Human Action Recognition from Skeleton Datacited by 105
  15. 2016Spatiotemporal Residual Networks for Video Action Recognitioncited by 91
  16. 2016Long-term Temporal Convolutions for Action Recognitioncited by 84
  17. 2018View Adaptive Neural Networks for High Performance Skeleton-based Human Action Recognitioncited by 84
  18. 2019Meta-Learning of Neural Architectures for Few-Shot Learningcited by 73
  19. 2019ProtoGAN: Towards Few Shot Learning for Action Recognitioncited by 41
  20. 2018A Generative Approach to Zero-Shot and Few-Shot Action Recognitioncited by 39
  21. 2019Spatiotemporal Pyramid Network for Video Action Recognitioncited by 37
  22. 2020Revisiting Pose-Normalization for Fine-Grained Few-Shot Recognitioncited by 31
  23. 2020TAFSSL: Task-Adaptive Feature Sub-Space Learning for few-shot classificationcited by 30
  24. 2019A Comparative Review of Recent Kinect-based Action Recognition Algorithmscited by 29
  25. 2020Tensor Representations for Action Recognitioncited by 28
  26. 2014Histogram of Oriented Principal Components for Cross-View Action Recognitioncited by 27
  27. 2020Skeleton-DML: Deep Metric Learning for Skeleton-Based One-Shot Action Recognitioncited by 23
  28. 2020SL-DML: Signal Level Deep Metric Learning for Multimodal One-Shot Action Recognitioncited by 18
  29. 2014HOPC: Histogram of Oriented Principal Components of 3D Pointclouds for Action Recognitioncited by 15
  30. 2018CNN-based Action Recognition and Supervised Domain Adaptation on 3D Body Skeletons via Kernel Feature Mapscited by 13
  31. 2018Power Normalizing Second-order Similarity Network for Few-shot Learningcited by 13
  32. 2016Tensor Representations via Kernel Linearization for Action Recognition from 3D Skeletons (Extended Version)cited by 10
  33. 2016Kernelized Covariance for Action Recognitioncited by 1

Built on by

Cited by year

Citing papers per year: 2022: 3; 2023: 2; 2024: 3; 2025: 2.

10 papers.

  1. 2024Advancing Video Anomaly Detection: A Concise Review and a New Datasetcited by 49
  2. 20233Mformer: Multi-order Multi-mode Transformer for Skeletal Action Recognitioncited by 40
  3. 2024Learnable Expansion of Graph Operators for Multi-Modal Feature Fusioncited by 23
  4. 2022Temporal-Viewpoint Transportation Plan for Skeletal Few-shot Action Recognitioncited by 22
  5. 2022Multi-level Second-order Few-shot Learningcited by 12
  6. 2023Traffic Forecasting on New Roads Using Spatial Contrastive Pre-Training (SCPT)cited by 11
  7. 2024Meet JEANIE: a Similarity Measure for 3D Skeleton Sequences via Temporal-Viewpoint Alignmentcited by 10
  8. 2025Feature Hallucination for Self-supervised Action Recognitioncited by 3
  9. 2025Few-Shot Learning in Video and 3D Object Detection: A Surveycited by 2
  10. 2022Part-aware Prototypical Graph Network for One-shot Skeleton-based Action Recognitioncited by 1