Research map / Video / Action recognition
Unsupervised Representation Learning With Long-Term Dynamics for Skeleton Based Action Recognition
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| pretrain with conditional skeleton inpainting (GAN) representation | supervised GRU baseline · CMU subset | accuracy: 87.83 -> 91.19 | Helped |
| pretrain with Unsupervised LSTM | supervised GRU baseline · CMU | accuracy: 79.89 -> 78.63 | Hurt |
| long-term inpainting objective instead of autoencoder reconstruction | unsupervised representation · CMU subset | accuracy: 77.03 -> 84.57 | Helped |
| add adversarial loss to element-wise loss (ablation) | inpainting model w/ F · CMU | classification accuracy: 62.56 -> 66.23 | Helped |
| add adversarial loss to element-wise loss (ablation) | inpainting model w/ F · CMU | inpainting MSE: 0.20 -> 0.16 | Helped |
| add adversarial loss (ablation) | inpainting model w/ F · CMU | clustering accuracy: 45.27 -> 51.35 | Helped |
| condition decoder on learned representation F (ablation) | inpainting model, joint loss · CMU | inpainting MSE: 0.23 -> 0.16 | Helped |
| unsupervised pretraining with only 2 labeled examples per class | supervised model · CMU | accuracy: 45.8% -> 49.73% | Helped |
Closest papers by meaning
- PREDICT & CLUSTER: Unsupervised Skeleton Based Action Recognition (CVPR 2020 · similarity 0.91)
- Heterogeneous Skeleton-Based Action Representation Learning (CVPR 2025 · similarity 0.89)
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.89)
- Recovering Complete Actions for Cross-dataset Skeleton Action Recognition (NeurIPS 2024 · similarity 0.88)
- Neuron: Learning Context-Aware Evolving Representations for Zero-Shot Skeleton Action Recognition (CVPR 2025 · similarity 0.88)
- Stitch, Contrast, and Segment: Learning a Human Action Segmentation Model Using Trimmed Skeleton Videos (AAAI 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.