Research map / Video / Action recognition
Time-Conditioned Action Anticipation in One Shot
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 the proposed time-conditioned method instead of TSN | Action anticipation on Epic-Kitchen · Epic-Kitchen, next action | Top-1 accuracy: 23.8% -> 25.6% | Helped |
| Use the proposed time-conditioned method instead of TSN | Action anticipation on Epic-Kitchen · Epic-Kitchen, next action | AvgCP: 2.9% -> 6.3% | Helped |
| Anticipate directly in one shot rather than iteratively (ablation) | Time-conditioned one-shot anticipation · Epic-Kitchen, long-term anticipation | anticipation accuracy | Helped |
| Anticipate directly in one shot rather than iteratively (ablation) | Time-conditioned one-shot anticipation · 50Salads, future actions within 10s | anticipation accuracy | Hurt |
| Add an attended temporal feature instead of averaging temporal features (ablation) | Time-conditioned one-shot anticipation · Epic-Kitchen and 50Salads | anticipation accuracy | Helped |
| Add an attended temporal feature instead of averaging temporal features (ablation) | Time-conditioned one-shot anticipation · Epic-Kitchen, long-term anticipation | anticipation accuracy | No clear effect |
| Add a skip connection from the last observed action without conditioning its weight on time (ablation) | TOS AF · Epic-Kitchen, future 60s | anticipation accuracy: 20.5% -> 15.5% | Hurt |
| Condition the skip-connection weight on time (ablation) | Attended-feature one-shot anticipation with skip connection · Epic-Kitchen, short- and long-term anticipation | anticipation accuracy | Helped |
| Use an attended temporal feature instead of only the last observed action (ablation) | TOS TSC · Epic-Kitchen, future 60s | anticipation accuracy: 2.3% | Helped |
| Add an attended temporal feature instead of using only the last observed action (ablation) | TOS TSC · Epic-Kitchen, long-term anticipation | anticipation accuracy | Helped |
| Include more than the last observation in the skip connection (ablation) | Proposed method · Epic-Kitchen, future 1s–50s | anticipation accuracy | No clear effect |
| Use the proposed method instead of RNN | Dense action anticipation · 50Salads, observation 20%, prediction 10% | mean over classes accuracy: 0.3006 -> 0.3251 | Helped |
Closest papers by meaning
- MANTA: Diffusion Mamba for Efficient and Effective Stochastic Long-Term Dense Action Anticipation (CVPR 2025 · similarity 0.87)
- Towards Understanding Future: Consistency Guided Probabilistic Modeling for Action Anticipation (AAAI 2024 · similarity 0.87)
- AntGPT: Can Large Language Models Help Long-term Action Anticipation from Videos? (ICLR 2024 · similarity 0.87)
- Can't Make an Omelette Without Breaking Some Eggs: Plausible Action Anticipation Using Large Video-Language Models (CVPR 2024 · similarity 0.86)
- Uncertainty-aware Action Decoupling Transformer for Action Anticipation (CVPR 2024 · similarity 0.85)
- ActFusion: a Unified Diffusion Model for Action Segmentation and Anticipation (NeurIPS 2024 · similarity 0.85)
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.