Research map / Detection and segmentation / Open-set and few-shot detection
Object-Oriented Dynamics Predictor
Guangxiang Zhu, Zhiao Huang, Chongjie Zhang
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 |
|---|---|---|---|
| use object-level masks and action-conditioned object relations rather than pixel-level dynamics prediction | OODP · Monster Kong unseen environments, 3-to-10 | 0-error accuracy: 0.20 -> 0.86 | Improved |
| use object-level masks and action-conditioned object relations rather than pixel-level dynamics prediction | OODP · Monster Kong unseen environments, 3-to-10 | 1-error accuracy: 0.47 -> 0.94 | Improved |
| use object-level masks and action-conditioned object relations rather than pixel-level dynamics prediction | OODP · Monster Kong unseen environments, 3-to-10 | 2-error accuracy: 0.56 -> 0.96 | Improved |
| add proposal loss for dynamic-region guidance (ablation) | OODP · Monster Kong unseen environments, 3-to-10 | 0-error accuracy: 0.86 -> 0.73 | Hurt |
| add proposal loss for dynamic-region guidance (ablation) | OODP · Monster Kong unseen environments, 2-to-10 | 1-error accuracy: 0.91 -> 0.90 | No clear effect |
| add proposal loss for dynamic-region guidance (ablation) | OODP · Monster Kong unseen environments, 5-to-10 | 2-error accuracy: 0.98 -> 0.99 | Improved |
| increase the maximum object-mask count from 3 to 5 (ablation) | OODP · unseen environments | dynamics prediction accuracy: 0.840 -> 0.842 | No clear effect |
Closest papers by meaning
- Learning Causal Dynamics Models in Object-Oriented Environments (ICML 2024 · similarity 0.87)
- QORA: Zero-Shot Transfer via Interpretable Object-Relational Model Learning (ICML 2024 · similarity 0.85)
- The Surprising Ineffectiveness of Pre-Trained Visual Representations for Model-Based Reinforcement Learning (NeurIPS 2024 · similarity 0.85)
- PlaySlot: Learning Inverse Latent Dynamics for Controllable Object-Centric Video Prediction and Planning (ICML 2025 · similarity 0.85)
- Efficient Exploration and Discriminative World Model Learning with an Object-Centric Abstraction (ICLR 2025 · similarity 0.85)
- Learning Interactive World Model for Object-Centric Reinforcement Learning (NeurIPS 2025 · similarity 0.85)
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.
