description: PhenoBench is a large-scale dataset for semantic image interpretation
  in the agricultural domain. It provides 2,872 high-resolution UAV images (1024x1024)
  of a real sugar beet field at three growth stages, with dense pixel-wise semantic
  segmentation (background/crop/weed/partial-crop/partial-weed), plant-level instance
  segmentation, and crop leaf-level instance segmentation, each with a per-instance
  visibility score. The test split is held out (unlabeled) for the official PhenoBench
  benchmark server.
format: FiftyOneDataset
license: cc-by-sa-4.0
name: PhenoBench
tags:
- aerial
- agriculture
- fiftyone
- image
- image-segmentation
- instance-segmentation
- leaf-segmentation
- object-detection
- panoptic-segmentation
- plant-phenotyping
- precision-agriculture
- semantic-segmentation
- sugar-beet
- uav
- weed-detection