Research map / Reinforcement learning / Visual and representation RL
Unsupervised Active Visual Search With Monte Carlo Planning Under Uncertain Detections
Francesco Taioli, Francesco Giuliari, Yiming Wang, Riccardo Berra, Alberto Castellini, Alessio Del Bue, Alessandro Farinelli, Marco Cristani, Francesco Setti
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 |
|---|---|---|---|
| add new belief update (exclude observed locations) (ablation) | POMP · AVDB, GT annotations, 8 scenes | SR: 0.67 -> 0.79 | Improved |
| add new belief update (ablation) | POMP · AVDB Medium scenes, GT | SR: 0.76 -> 0.85 | Improved |
| add new belief update (ablation) | POMP · AVDB Hard scenes, GT | SR: 0.48 -> 0.70 | Improved |
| add new belief update (ablation) | POMP · AVDB Easy scenes, GT | SR / APL: SR 0.84 -> 0.83; APL 18.31 -> 14.65 | No clear effect |
| add new belief update (ablation) | POMP · AVDB, real detector, 8 scenes | SR: 0.40 -> 0.44 | Improved |
| add probabilistic detection and docking (ablation) | POMP · AVDB, real detector, 8 scenes | SR: 0.40 -> 0.49 | Improved |
| add probabilistic detection and docking (ablation) | POMP · AVDB, real detector, 8 scenes | SPL: 0.50 -> 0.40 | Hurt |
| add probabilistic detection and docking (ablation) | POMP · AVDB, real detector, 8 scenes | APL: 41.13 -> 48.92 | Hurt |
| add probabilistic detection to POMP (false-positive terminations) (ablation) | POMP · AVDB, real detector | false-positive localisation errors: ~32% decrease | Improved |
| use probabilistic docking (knowledge from planning) instead of detector-guided docking (ablation) | POMP · AVDB, real detector | docking error: ~35.7% decrease | Improved |
| add new belief update on top of probabilistic detection (ablation) | POMP-PD · AVDB, real detector, 8 scenes | SPL: 0.40 -> 0.47 | Improved |
| POMP-BE-PD versus POMP with a real detector | POMP · AVDB, real detector, 8 scenes | SR: 0.40 -> 0.54 | Improved |
Closest papers by meaning
- POMDP Planning for Object Search in Partially Unknown Environment (NeurIPS 2023 · similarity 0.88)
- Evidential Active Recognition: Intelligent and Prudent Open-World Embodied Perception (CVPR 2024 · similarity 0.84)
- Simplifying Complex Observation Models in Continuous POMDP Planning with Probabilistic Guarantees and Practice (AAAI 2024 · similarity 0.84)
- Reasoning in Visual Navigation of End-to-end Trained Agents: A Dynamical Systems Approach (CVPR 2025 · similarity 0.84)
- OVOD-Agent: A Markov-Bandit Framework for Proactive Visual Reasoning and Self-Evolving Detection (CVPR 2026 · similarity 0.84)
- BeliefMapNav: 3D Voxel-Based Belief Map for Zero-Shot Object Navigation (NeurIPS 2025 · similarity 0.84)
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.
