Research map / Video / Action recognition
Modular Action Concept Grounding in Semantic Video Prediction
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 |
|---|---|---|---|
| Use language-guided concept slots instead of concatenated features and a tiled action vector (ablation) | MAC · CLEVR-Building-blocks | Accuracy: 50.8% -> 95.2% | Helped |
| Use language-guided concept slots instead of concatenated features and a tiled action vector (ablation) | MAC · Sapien-Kitchen | Accuracy: 52.4% -> 86.4% | Helped |
| Use language-guided concept slots instead of concatenated features and a tiled action vector (ablation) | MAC · CLEVR-Building-blocks | MSE: 226.53 -> 43.52 | Helped |
| Use language-guided concept slots instead of concatenated features and a tiled action vector (ablation) | MAC · Sapien-Kitchen | MSE: 23.13 -> 11.16 | Helped |
| Use MAC rather than AG2Vid, which requires bounding boxes | semantic action-conditional video prediction · CLEVR-Building-blocks | Accuracy: 78.8% -> 95.2% | Helped |
| Use MAC rather than AG2Vid, which requires bounding boxes | semantic action-conditional video prediction · Sapien-Kitchen | Accuracy: 5.2% -> 86.4% | Helped |
| Use stochastic MAC with concept slots rather than action-conditional SVG-LP | stochastic video prediction · Tower-Creation | Accuracy: 31.8% -> 65.3% | Helped |
| Train without red-cube videos, then test instructions involving red cubes | MAC · CLEVR-Building-blocks, unobserved red-cube scenarios | Accuracy: 75.6% | Helped |
Closest papers by meaning
- Relational Action Forecasting (CVPR 2019 · similarity 0.89)
- VideoNet: A Large-Scale Dataset for Domain-Specific Action Recognition (CVPR 2026 · similarity 0.88)
- Large-Scale Weakly-Supervised Pre-Training for Video Action Recognition (CVPR 2019 · similarity 0.88)
- SeFAR: Semi-supervised Fine-grained Action Recognition with Temporal Perturbation and Learning Stabilization (AAAI 2025 · similarity 0.87)
- Disentangled Concepts Speak Louder Than Words: Explainable Video Action Recognition (NeurIPS 2025 · similarity 0.86)
- Open-Vocabulary Video Relation Extraction (AAAI 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.