Research map / Video / Action recognition
Procedure Planning in Instructional Videos
Chien-Yi Chang, De-An Huang, Danfei Xu, Ehsan Adeli, Li Fei-Fei, Juan Carlos Niebles
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 conjugate dynamics P to forward dynamics T (ablation) | DDN · CrossTask procedure planning (T=3) | Success Rate: <0.01% -> 12.18% | Improved |
| add conjugate dynamics P to forward dynamics T (ablation) | DDN · CrossTask procedure planning (T=4) | Accuracy: 2.51% -> 27.10% | Improved |
| add forward dynamics T to conjugate dynamics P (ablation) | DDN · CrossTask procedure planning (T=3) | Success Rate: 1.55% -> 12.18% | Improved |
| add forward dynamics T to conjugate dynamics P (ablation) | DDN · CrossTask procedure planning (T=4) | mIoU: 26.54% -> 48.46% | Improved |
| add forward dynamics T to conjugate dynamics P (ablation) | DDN · CrossTask procedure planning (T=3) | Accuracy: 18.66% -> 31.29% | Improved |
| add conjugate dynamics P to forward dynamics T (ablation) | DDN · CrossTask walkthrough planning (T=3) | Pairwise Accuracy: 83.33% -> 86.81% | Improved |
| add forward dynamics T to conjugate dynamics P (ablation) | DDN · CrossTask walkthrough planning (T=3) | Pairwise Accuracy: 50.45% -> 86.81% | Improved |
| add forward dynamics T to conjugate dynamics P (ablation) | DDN · CrossTask walkthrough planning (T=4) | Hamming Distance: 2.01 -> 0.88 | Improved |
| use only P (w/o T) rather than only T (w/o P) (ablation) | DDN ablations · CrossTask walkthrough planning (T=3) | Pairwise Accuracy: 83.33% -> 50.45% | Hurt |
| replace gradient-based UPN planner with DDN | UPN · CrossTask procedure planning (T=3) | Success Rate: 2.89% -> 12.18% | Improved |
| use DDN instead of retrieval baseline RB | RB · CrossTask procedure planning (T=4) | Success Rate: 3.95% -> 5.97% | Improved |
| use DDN instead of Causal InfoGAN | Causal InfoGAN · CrossTask walkthrough planning (T=3) | Pairwise Accuracy: 71.55% -> 86.81% | Improved |
Closest papers by meaning
- Procedure Knowledge Decoupled Distillation Strategy for Procedure Planning in Instructional Videos (AAAI 2025 · similarity 0.90)
- Masked Temporal Interpolation Diffusion for Procedure Planning in Instructional Videos (ICLR 2025 · similarity 0.89)
- PlanLLM: Video Procedure Planning with Refinable Large Language Models (AAAI 2025 · similarity 0.88)
- Learning an Actionable Discrete Diffusion Policy via Large-Scale Actionless Video Pre-Training (NeurIPS 2024 · similarity 0.87)
- Differentiable Learning of Lifted Action Schemas for Classical Planning (NeurIPS 2026 · similarity 0.87)
- Why Not Use Your Textbook? Knowledge-Enhanced Procedure Planning of Instructional Videos (CVPR 2024 · similarity 0.87)
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.
