ObjView-Bench: Rethinking Difficulty and Deployment for Object-Centric View Planning

2026-05-11

AI Summary

Existing evaluation protocols for object-centric view planning conflate object complexity, planning difficulty, budget assumptions, and physical reachability, obscuring true reconstruction performance. This work proposes ObjView-Bench, a novel benchmark that decouples view planning difficulty into three quantifiable dimensions: self-occlusion, observation saturation, and protocol dependency. The framework introduces a controlled dataset and a deployment-oriented evaluation protocol that jointly accounts for budget and reachability constraints. Within this unified setting, systematic experiments across classical, learning-based, and hybrid planners reveal that difficulty level, budget allocation, and reachability significantly influence method rankings and failure modes. Building on these insights, the study further introduces a difficulty-aware sampling strategy that effectively enhances the performance of learning-based planners.

Abstract

Object-centric view planning is a core component of active geometric 3D reconstruction in robotics, yet existing evaluations often conflate object complexity, planning difficulty, budget assumptions, and physical reachability constraints. As a result, conclusions drawn from idealized view-planning evaluations may not reliably predict performance under realistic reconstruction settings. We introduce ObjView-Bench, an evaluation framework for rethinking difficulty and deployment in object-centric view planning. First, we disentangle three quantities underlying view-planning evaluation: omnidirectional self-occlusion as an object-side attribute, observation saturation difficulty, and protocol-dependent planning difficulty defined through a set-cover formulation. This separation supports controlled dataset construction, analysis of slow-saturation objects, and a case study showing that planning difficulty-aware sampling can improve learned view planners. Second, we design deployment-oriented evaluation protocols that reveal how budget regimes and reachable-view constraints alter method behavior. Across classical, learned, and hybrid planners, ObjView-Bench shows that difficulty, budget, and reachability constraints substantially change method rankings and failure modes.

Problem

• object-centric view planning
• evaluation framework
• planning difficulty
• 3D reconstruction
• reachability constraints

Innovation

• view planning
• object-centric reconstruction
• evaluation benchmark
• self-occlusion
• deployment constraints