Research map / Reinforcement learning / Visual and representation RL
Tree-Structured Reinforcement Learning for Sequential Object Localization
Read the paperCheck 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 |
|---|---|---|---|
| Branch into the best scaling and translation actions instead of following only one highest-value action (ablation) | Tree-RL · PASCAL VOC 2007 test, all objects, IoU=0.5, 31 proposals | recall (%): 53.8 -> 68.1 | Helped |
| Branch into the best scaling and translation actions instead of following only one highest-value action (ablation) | Tree-RL · PASCAL VOC 2007 test, small objects, IoU=0.6, 31 proposals | recall (%): 15.6 -> 12.5 | Hurt |
| Increase the tree-search budget from 31 to 63 proposals (ablation) | Tree-RL · PASCAL VOC 2007 test, all objects, IoU=0.5 | recall (%): 68.1 -> 74.8 | Helped |
| Use Tree-RL proposals instead of RPN proposals with the same Fast R-CNN detector | Fast R-CNN (ResNet-101) · PASCAL VOC 2007 test | mAP: 75.8 -> 76.6 | Helped |
| Use Tree-RL proposals instead of RPN proposals with the same Fast R-CNN detector | Fast R-CNN (ResNet-101) · PASCAL VOC 2012 test | mAP: 73.1 -> 73.7 | Helped |
| Use Tree-RL proposals with Fast R-CNN instead of Faster R-CNN | Object detection on PASCAL VOC 2007 · PASCAL VOC 2007 test | mAP: 76.4 -> 76.6 | Helped |
| Use Tree-RL proposals with Fast R-CNN instead of Faster R-CNN | Object detection on PASCAL VOC 2012 · PASCAL VOC 2012 test | mAP: 73.8 -> 73.7 | Hurt |
Closest papers by meaning
- RELO: Reinforcement Learning to Localize for Visual Object Tracking (ICML 2026 · similarity 0.88)
- Structured World Belief for Reinforcement Learning in POMDP (ICML 2021 · similarity 0.87)
- Reinforcement Learning for Visual Object Detection (CVPR 2016 · similarity 0.87)
- Efficient Exploration and Discriminative World Model Learning with an Object-Centric Abstraction (ICLR 2025 · similarity 0.86)
- Entity-Centric Reinforcement Learning for Object Manipulation from Pixels (ICLR 2024 · similarity 0.85)
- Tree-Guided Diffusion Planner (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.