Research map / Vision-language and multimodal / Human and embodied modeling
Predicting Motivations of Actions by Leveraging Text
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 |
|---|---|---|---|
| Add text-mined language potentials to visual classifiers | Vision model · Motivations dataset test | median rank (motivation): 39 -> 28 | Helped |
| Add the person crop features | Vision model · Motivations dataset test | median rank (motivation): 39 -> 42 | Hurt |
| Add person crop features to the text-augmented model | Vision + Text · Motivations dataset test | median rank (motivation): 28 -> 27 | Helped |
| Use a web-trained language model instead of the Wikipedia language model | Vision + Person + Text · Motivations dataset test | median rank (motivation): 28 -> 27 | Helped |
| Provide the ground-truth action to the model (ablation) | Vision + Person + Text · Motivations dataset test | median rank (motivation): 27 -> 26 | Helped |
| Provide the ground-truth scene to the model (ablation) | Vision + Person + Text · Motivations dataset test | median rank (motivation): 27 -> 27 | No clear effect |
| Provide both the ground-truth action and scene to the model (ablation) | Vision + Person + Text · Motivations dataset test | median rank (motivation): 27 -> 25 | Helped |
| Add text-mined language potentials to visual classifiers | Vision model · Motivations dataset test | median rank (action): 17 -> 15 | Helped |
| Add text-mined language potentials to visual classifiers | Vision model · Motivations dataset test | median rank (scene): 4 -> 3 | Helped |
Closest papers by meaning
- Intentonomy: A Dataset and Study Towards Human Intent Understanding (CVPR 2021 · similarity 0.87)
- Reward Learning From Narrated Demonstrations (CVPR 2018 · similarity 0.85)
- Learning Goals From Failure (CVPR 2021 · similarity 0.85)
- Bridge-Prompt: Towards Ordinal Action Understanding in Instructional Videos (CVPR 2022 · similarity 0.85)
- LLaVAction: evaluating and training multi-modal large language models for action understanding (ICLR 2026 · similarity 0.83)
- Disentangled Concepts Speak Louder Than Words: Explainable Video Action Recognition (NeurIPS 2025 · 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.