Research map / Video / Action recognition
Don't Pour Cereal into Coffee: Differentiable Temporal Logic for Temporal Action Segmentation
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 |
|---|---|---|---|
| add DTL temporal logic loss | GRU · 50Salads | Edit: 55.2 -> 62.1 | Helped |
| add DTL temporal logic loss | MS-TCN · 50Salads | F1@50: 64.5 -> 67.6 | Helped |
| add DTL temporal logic loss | ASFormer · 50Salads | Edit: 76.9 -> 80.5 | Helped |
| add DTL temporal logic loss | GRU · Breakfast | F1@50: 38.4 -> 40.7 | Helped |
| add DTL temporal logic loss | MS-TCN · Breakfast | Edit: 71.2 -> 71.6 | Helped |
| add DTL temporal logic loss | ASFormer · Breakfast | F1@50: 60.5 -> 62.9 | Helped |
| add DTL temporal logic loss | all models · 50Salads and Breakfast | frame-wise accuracy: 79.0 -> 80.3 | No clear effect |
| apply only before/during (BD) constraints (ablation) | GRU · 50Salads | F1@10: -1.3 | Hurt |
| apply only FC constraints (ablation) | GRU · 50Salads | Edit: 7.0 | Helped |
| apply only Ip constraints (ablation) | GRU · 50Salads | Edit: -0.8 | Hurt |
| apply only Ip constraints (ablation) | MS-TCN · 50Salads | F1@25: 2.8 | Helped |
| apply all constraints versus individual (ablation) | GRU · 50Salads | F1@25: 6.6 -> 6.0 | Hurt |
Closest papers by meaning
- Enhancing Temporal Understanding in Video-LLMs through Stacked Temporal Attention in Vision Encoders (NeurIPS 2025 · similarity 0.84)
- FACT: Frame-Action Cross-Attention Temporal Modeling for Efficient Action Segmentation (CVPR 2024 · similarity 0.83)
- Masked Diffusion Vision-Language Models for Temporal Action Localization (NeurIPS 2026 · similarity 0.83)
- OnlineTAS: An Online Baseline for Temporal Action Segmentation (NeurIPS 2024 · similarity 0.83)
- EventLens: Event-Structure Reinforcement Learning for Video Understanding (NeurIPS 2026 · similarity 0.83)
- ReXTime: A Benchmark Suite for Reasoning-Across-Time in Videos (NeurIPS 2024 · similarity 0.83)
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.