WildRoadBench - Drainage Damage: leaderboard

The drainage damage class of WildRoadBench (2026), an aerial road-damage grounding benchmark: models box every instance in wild drone images, 22 ground-truth boxes in all. The score is F1 at IoU 0.5 in percent; higher is better.

Metric: F1 (%) of predicted boxes at IoU 0.5 for the drainage damage scene class (22 ground-truth boxes), VLM Track, zero-shot box grounding on the full 1,061-image aerial set (1,394 ground-truth boxes); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 25 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Qwen 3.6 Plus58.3#245
2Gemini 3 Pro57.8#146
3Gemini 3.1 Flash Lite50#377
4Qwen3.5-VL-35B-A3B-FP850
5Qwen 3 VL 8B Instruct45.5#1043
6Qwen3-VL-32B-Instruct-AWQ45.5
7Qwen3-VL-32B-Thinking-AWQ45.5
8Qwen 3.5 VL 9B43.5
9Claude Sonnet 4.642.3#179
10Qwen3.5-VL-27B-FP836.4
11Kimi K2.535.1#272
12Qwen2.5-VL-32B-Instruct-AWQ31.8
13Qwen 3.5 VL 4B31.1
14Qwen 2.5 VL 7B Instruct23.8#1919
15Gemini 2.5 Flash22.7#466

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=wildroadbench-drainage-damage · How It Works · Data refreshed daily, snapshot 2026-11-22.