BigEarthNet.txt - Referring LULC Detection: leaderboard
Metric: Mean IoU (%) of the bounding box predicted for a land-use/land-cover instance described in text, without a location hint, of the BigEarthNet.txt benchmark split (1,082 manually verified Sentinel-1/Sentinel-2 image pairs over Europe from the BigEarthNet v2.0 test set), zero-shot, each model given the RGB bands of the Sentinel-2 image with the benchmark instruction and answers extracted even when off-format; higher is better. Source: arxiv.org. Saturation forecast: Around March 2028. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.2 | 25.16 | #105 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=bigearthnet-txt-referring-lulc-detection · How It Works · Data refreshed daily, snapshot 2026-10-11.