Research map / Video / Action recognition
Learning the Predictability of the Future
Check 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 |
|---|---|---|---|
| Use hyperbolic rather than Euclidean representations at 256 dimensions | Early action prediction · FineGym | accuracy (%): 68.16 -> 82.54 | Helped |
| Use hyperbolic rather than Euclidean representations at 256 dimensions | Early action prediction · Hollywood2 | accuracy: 21.77 -> 23.10 | Helped |
| Use hyperbolic rather than Euclidean representations at 256 dimensions | Early action prediction · Hollywood2 | top-down hierarchical accuracy: 33.00 -> 33.99 | Helped |
| Use hyperbolic rather than Euclidean representations at 256 dimensions | Early action prediction · Hollywood2 | bottom-up hierarchical accuracy: 27.38 -> 28.55 | Helped |
| Use hyperbolic rather than Euclidean representations at 256 dimensions | Future action prediction · FineGym | accuracy: 10.08 -> 13.37 | Helped |
| Use hyperbolic rather than Euclidean representations at 256 dimensions | Future action prediction · FineGym | top-down hierarchical accuracy: 52.00 -> 66.64 | Helped |
| Use hyperbolic rather than Euclidean representations at 256 dimensions | Future action prediction · FineGym | bottom-up hierarchical accuracy: 24.75 -> 33.04 | Helped |
| Reduce representation size from 256 to 64 dimensions (ablation) | Hyperbolic early action prediction · FineGym | accuracy (%): 82.54 -> 73.35 | Hurt |
| Reduce representation size from 256 to 64 dimensions (ablation) | Hyperbolic early action prediction · Hollywood2 | accuracy: 23.10 -> 22.25 | Hurt |
| Reduce representation size from 256 to 64 dimensions (ablation) | Hyperbolic future action prediction · FineGym | top-down hierarchical accuracy: 66.64 -> 56.67 | Hurt |
| Reduce representation size from 256 to 64 dimensions (ablation) | Hyperbolic future action prediction · FineGym | accuracy: 13.37 -> 10.29 | Hurt |
| Reduce representation size from 256 to 64 dimensions (ablation) | Hyperbolic early action prediction · Hollywood2 | bottom-up hierarchical accuracy: 28.55 -> 26.86 | Hurt |
Closest papers by meaning
- Anticipating Visual Representations From Unlabeled Video (CVPR 2016 · similarity 0.90)
- PlaySlot: Learning Inverse Latent Dynamics for Controllable Object-Centric Video Prediction and Planning (ICML 2025 · similarity 0.89)
- Learning to Generate Long-term Future via Hierarchical Prediction (ICML 2017 · similarity 0.88)
- Video Prediction via Example Guidance (ICML 2020 · similarity 0.88)
- FutureHuman3D: Forecasting Complex Long-Term 3D Human Behavior from Video Observations (CVPR 2024 · similarity 0.86)
- Can't Make an Omelette Without Breaking Some Eggs: Plausible Action Anticipation Using Large Video-Language Models (CVPR 2024 · similarity 0.86)
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.