Research map / Multi-agent systems and games / Human-AI and goal-conditioned
Hierarchical Object-to-Zone Graph for Object Navigation
Sixian Zhang, Xinhang Song, Yubing Bai, Weijie Li, Yakui Chu, Shuqiang Jiang
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 |
|---|---|---|---|
| train and test with sub-goal zone instead of target zone (ablation) | HOZ · AI2-THOR (All) | SR: 69.16 -> 70.57 | Improved |
| add the object layer to the baseline (ablation) | A3C baseline · AI2-THOR (All) | SR: 57.35 -> 65.12 | Improved |
| add the scene layer on top of the object layer (ablation) | A3C + object layer · AI2-THOR (All) | SPL: 37.86 -> 38.83 | Improved |
| add the zone layer on top of object and scene layers (ablation) | A3C + object + scene · AI2-THOR (All) | SR: 65.81 -> 70.57 | Improved |
| add the done reminder to the object-layer baseline (ablation) | A3C + object layer · AI2-THOR (All) | SR: 65.12 -> 66.73 | Improved |
| add the done reminder to full HOZ (ablation) | HOZ · AI2-THOR (All) | SPL: 40.84 -> 40.02 | Hurt |
| add self-supervised test-time adaptation (TPN) to HOZ (ablation) | HOZ · AI2-THOR (All) | SR: 70.62 -> 73.15 | Mixed |
| add location information to visual features when clustering zones (ablation) | HOZ · AI2-THOR (All) | SR: -2.40 | Hurt |
| use ground-truth object detection instead of pretrained detector (ablation) | A3C + object layer · AI2-THOR (All) | SPL: 37.86 -> 55.91 | Improved |
| use ground-truth object detection instead of pretrained detector (ablation) | A3C + object layer · AI2-THOR (All) | SR: 65.12 -> 66.78 | No clear effect |
| use learned zone nodes instead of random nodes (edges random) (ablation) | HOZ · AI2-THOR (All) | SR: 67.81 -> 69.33 | Improved |
| use learned edges instead of random edges (nodes kept) (ablation) | HOZ · AI2-THOR (All) | SR: 69.33 -> 70.47 | Improved |
Closest papers by meaning
- HOZ++: Versatile Hierarchical Object-to-Zone Graph for Object Navigation (TPAMI 2025 · similarity 0.92)
- Trajectory Diffusion for ObjectGoal Navigation (NeurIPS 2024 · similarity 0.86)
- Hierarchical Semantic-Augmented Navigation: Optimal Transport and Graph-Driven Reasoning for Vision-Language Navigation (NeurIPS 2025 · similarity 0.86)
- SG-Nav: Online 3D Scene Graph Prompting for LLM-based Zero-shot Object Navigation (NeurIPS 2024 · similarity 0.85)
- Affordances-Oriented Planning Using Foundation Models for Continuous Vision-Language Navigation (AAAI 2025 · similarity 0.85)
- Efficient Exploration and Discriminative World Model Learning with an Object-Centric Abstraction (ICLR 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.
