Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Zero-Shot Anticipation for Instructional Activities
Fadime Sener, Angela Yao
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 Recipe1M text pre-training (100% vs none) (ablation) | Ours Text · Tasty Videos | BLEU4: 0.24 -> 3.30 | Improved |
| add Recipe1M text pre-training (100% vs none) (ablation) | Ours Text · Tasty Videos | METEOR: 6.80 -> 17.97 | Improved |
| increase pre-training data from 50% to 100% (ablation) | Ours Text · Tasty Videos | BLEU4: 2.42 -> 3.30 | Improved |
| add ingredient input (ablation) | Ours Text · Tasty Videos | METEOR: 13.07 -> 17.97 | Improved |
| add ingredient input (ablation) | Ours Text · Tasty Videos | BLEU4: 0.92 -> 3.30 | Improved |
| add Recipe1M pre-training (ablation) | Sup. Visual · YouCookII cross-validation | METEOR: 5.89 -> 10.22 | Improved |
| add Recipe1M pre-training (ablation) | Zero Visual · YouCookII cross-validation | METEOR: 5.03 -> 9.27 | Improved |
| train on target dishes (supervised instead of zero-shot) (ablation) | Visual model · YouCookII cross-validation | METEOR: 9.27 -> 10.22 | Improved |
| use ground-truth segments instead of fixed windows | Ours Visual · Tasty Videos | METEOR: 11.00 -> 11.78 | Improved |
| add joint sentence and video context to video model | Ours Visual · Tasty Videos | METEOR: 11.00 -> 14.09 | Improved |
| use text input instead of video | Ours · Tasty Videos | METEOR: 11.00 -> 17.97 | Improved |
| test on recipes with training variations (ablation) | Ours Visual · Tasty Videos | BLEU4: 0.76 -> 2.16 | Improved |
Closest papers by meaning
- Transferring Knowledge From Text to Video: Zero-Shot Anticipation for Procedural Actions (TPAMI 2022 · similarity 0.96)
- Can't Make an Omelette Without Breaking Some Eggs: Plausible Action Anticipation Using Large Video-Language Models (CVPR 2024 · similarity 0.89)
- RoboDreamer: Learning Compositional World Models for Robot Imagination (ICML 2024 · similarity 0.87)
- Zero-Shot Robotic Manipulation with Pre-Trained Image-Editing Diffusion Models (ICLR 2024 · similarity 0.86)
- AntGPT: Can Large Language Models Help Long-term Action Anticipation from Videos? (ICLR 2024 · similarity 0.86)
- Seer: Language Instructed Video Prediction with Latent Diffusion Models (ICLR 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.
