SelvaBox: A high-resolution dataset for tropical tree crown detection

TL;DR

This work introduces SelvaBox, the largest open-access dataset for tropical tree crown detection in high-resolution drone imagery and reveals two key findings: higher-resolution inputs consistently boost detection accuracy; and models trained exclusively on SelvaBox achieve competitive zero-shot detection performance on unseen tropical tree crown datasets, matching or exceeding competing methods.

Abstract

Detecting individual tree crowns in tropical forests is essential to study these complex and crucial ecosystems impacted by human interventions and climate change. However, tropical crowns vary widely in size, structure, and pattern and are largely overlapping and intertwined, requiring advanced remote sensing methods applied to high-resolution imagery. Despite growing interest in tropical tree crown detection, annotated datasets remain scarce, hindering robust model development. We introduce SelvaBox, the largest open-access dataset for tropical tree crown detection in high-resolution drone imagery. It spans three countries and contains more than 83,000 manually labeled crowns - an order of magnitude larger than all previous tropical forest datasets combined. Extensive benchmarks on SelvaBox reveal two key findings: (1) higher-resolution inputs consistently boost detection accuracy; and (2) models trained exclusively on SelvaBox achieve competitive zero-shot detection performance on unseen tropical tree crown datasets, matching or exceeding competing methods. Furthermore, jointly training on SelvaBox and three other datasets at resolutions from 3 to 10 cm per pixel within a unified multi-resolution pipeline yields a detector ranking first or second across all evaluated datasets. Our dataset, code, and pre-trained weights are made public.

Figures (19)

• Figure 1: The SelvaBox dataset. The illustrated samples are extracted from rasters recorded in Panama, Brazil and Ecuador with a spatial extent of $80\text{m}\times80\text{m}$ and a resolution of 1.2 to 5.1 cm per pixel. The red square on the right highlights a zoom of the Ecuador sample with a spatial extent of $40\text{m}\times40\text{m}$ at the same resolution.
• Figure 2: Distribution of box annotations size in SelvaBox per country.
• Figure 3: Multi-resolution vs. single-resolution on SelvaBox. RF1$_{75}$ for the best single-resolution methods from Tab. \ref{['table:benchmark_80m']} trained at fixed $80\times 80$ m extent vs multi-resolution approaches with varying crop augmentation ranges $[36,88]$, $[30,100]$, $[30,120]$. All methods are 'DINO 5-scale Swin L-384'.
• Figure 4: RF1 vs IoU threshold on $\textsc{SelvaBox}$. Comparison of two of our DINO-Swin-L variants and competing methods at different IoU thresholds. In this work we focus on RF1$_{75}$ (IoU 75). For each IoU threshold, NMS hyperparameters are independently optimized on the validation set. Results for other datasets are in Appendix \ref{['section:rf1_iou_ablation']}.
• Figure 5: Visualization of spatially separated splits. All 14 rasters of SelvaBox are illustrated with their corresponding train, valid and test AOI-based splits. Images are uniformly sized and not at scale. A few train AOIs (red) have holes to exclude sparse annotations (see Section \ref{['sec:dataset_description']}).