Benchmark RadarAI BENCHMARK PROFILE
OV3D-Bench
OV3D-Bench is a diagnostic benchmark for open-vocabulary monocular 3D detection, evaluating localization, semantic robustness, and cross-domain transfer across seven datasets.
- Released
- 2026-08-17
- Readiness
- Runnable
- Primary field
- General AI
Why it matters
It reveals that open-vocabulary semantics is the primary bottleneck, providing a decoupled evaluation axis for detector diagnosis.
Motivation
Open-vocabulary monocular 3D detectors report strong in-domain performance, but each evaluates under a different protocol, several rely on per-image category oracles unavailable at deployment, and all collapse geometry and semantics into a single AP metric.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.