Research map / Video / Action recognition
ESTANet: Efficient Online Error Detection in Procedural Videos via Prediction Inconsistency
Shih-Po Lee, Reza Ghoddoosian, Faizan Siddiqui, Enna Sachdeva, Behzad Dariush
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 the TAD module to the detectors (ablation) | ESTANet · EgoPER | F1@50: 7.5 -> 21.6 | Improved |
| add the TAD module to the detectors (ablation) | ESTANet · Assembly-101-O | Avg-F1: 55.1 -> 59.6 | Improved |
| add the TAD module to the detectors (ablation) | ESTANet · EPIC-Tent-O | Avg-F1: 69.3 -> 70.2 | Improved |
| choose l by covering beta=2 preceding actions instead of 2nd shortest action duration (ablation) | ESTANet · EPIC-Tent-O | Avg-F1: 52.1 -> 70.2 | Improved |
| choose l by coverage strategy instead of longest action duration (ablation) | ESTANet · Assembly-101-O | Avg-F1: 45.3 -> 59.6 | Improved |
| choose l by coverage strategy instead of 2nd shortest action duration (ablation) | ESTANet · EgoPER | F1@50: 15.9 -> 21.6 | Improved |
| vary beta from 1 to 3 (ablation) | ESTANet · EPIC-Tent-O | Avg-F1: 66.6 / 70.2 / 65.6 (beta 1/2/3) | Mixed |
| use theta=50 instead of 20 (ablation) | ESTANet · Assembly-101-O | Avg-F1: 42.8 -> 59.6 | Improved |
| use theta=50 instead of 80 (ablation) | ESTANet · Assembly-101-O | Avg-F1: 44.3 -> 59.6 | Improved |
| use theta=50 instead of 20 or 80 (ablation) | ESTANet · EgoPER | F1@50: 15.6 / 17.4 -> 21.6 | Improved |
| replace DTGL with ESTANet | online error detection · EgoPER | F1@10: 39.6 -> 47.2 | Improved |
| replace DTGL with ESTANet | online error detection · EPIC-Tent-O | C-F1: 69.3 -> 48.0 | Hurt |
Closest papers by meaning
- Modeling Multiple Normal Action Representations for Error Detection in Procedural Tasks (CVPR 2025 · similarity 0.87)
- PREGO: Online Mistake Detection in PRocedural EGOcentric Videos (CVPR 2024 · similarity 0.87)
- Error Detection in Egocentric Procedural Task Videos (CVPR 2024 · similarity 0.86)
- Progress-Aware Online Action Segmentation for Egocentric Procedural Task Videos (CVPR 2024 · similarity 0.85)
- Differentiable Task Graph Learning: Procedural Activity Representation and Online Mistake Detection from Egocentric Videos (NeurIPS 2024 · similarity 0.84)
- CaptainCook4D: A Dataset for Understanding Errors in Procedural Activities (NeurIPS 2024 · similarity 0.84)
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.
