UniRef-UAV (Text-only): leaderboard

Metric: Precision@(F1=1, IoU>=0.5) (%; share of the 25,849 text-only test queries whose predicted box set exactly matches the referred target set under one-to-one IoU>=0.5 matching, an empty prediction being correct only for no-target queries; zero-shot, model-specific grounding prompt, unparseable output read as an empty prediction). Source: arxiv.org. Saturation forecast: Around May 2028. 14 models tracked.

Top models

#ModelScore
1Qwen 3.6 27B20.5
2Gemma 4 31B (IT)20.4

Interactive version: theaggregate.ai/benchmark?slug=uniref-uav-text-only · How It Works · Data refreshed daily, snapshot 2026-09-29.