# deadtrees.earth-aerial: A Multi-Resolution Aerial Image Dataset for Tree Cover and Mortality Detection

arXiv: https://arxiv.org/abs/2605.19605 · HTML: https://mlsanity.com/paper/2605.19605/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 40 references that are in the map.
- First posted on arXiv 19 May 2026.

## Builds on (most cited first, 10 of 10)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation](https://mlsanity.com/paper/1802.02611/) (2018)
- [Masked-attention Mask Transformer for Universal Image Segmentation](https://mlsanity.com/paper/2112.01527/) (2021)
- [Quantifying the Carbon Emissions of Machine Learning](https://mlsanity.com/paper/1910.09700/) (2019)
- [A Novel Focal Tversky loss function with improved Attention U-Net for lesion segmentation](https://mlsanity.com/paper/1810.07842/) (2018)
- [OAM-TCD: A globally diverse dataset of high-resolution tree cover maps](https://mlsanity.com/paper/2407.11743/) (2024)
- [SelvaMask: Segmenting Trees in Tropical Forests and Beyond](https://mlsanity.com/paper/2602.02426/) (2026)
- [Low-Cost Tree Crown Dieback Estimation Using Deep Learning-Based Segmentation](https://mlsanity.com/paper/2409.08171/) (2024)
- [Manual Labelling Artificially Inflates Deep Learning-Based Segmentation Performance on RGB Images of Closed Canopy: Validation Using TLS](https://mlsanity.com/paper/2503.14273/) (2025)

## Abstract (arXiv)

Forests worldwide are increasingly threatened by climate change and disturbances such as fire, pests, and pathogens, creating a need for scalable monitoring of tree cover and tree mortality. Aerial imagery from drones and aircraft enables detailed mapping of tree crowns and mortality, but progress is limited by the lack of globally distributed, multi-biome, harmonized datasets for joint segmentation of tree cover and mortality. To address this, we introduce two open, machine-learning-ready datasets for joint segmentation of tree cover and tree mortality from centimeter-scale aerial imagery across globally distributed sites. Our training dataset, DTE-aerial-train, comprises 346K image patches of 1024 × 1024 pixels at spatial resolutions from 2.5 to 20,cm, with audited pseudo-labels for tree cover and mortality and expert annotations for rare mortality sites. Our evaluation dataset, DTE-aerial-bench, is a multi-biome, multi-resolution benchmark of 25 globally distributed orthoimages (525 patches) with expert annotations for both tree cover and mortality. Both datasets span tropical, temperate, boreal, and dryland biomes and cover a wide range of forest structures and mortality patterns. Using the benchmark, we establish reference baselines for mortality and tree cover segmentation across biomes and spatial resolutions. Improvements over prior models vary by biome: they are largest in boreal and montane forests (F1 from 0.40 to 0.58) and smaller in tropical forests and drylands. For data, model and code, please visit our hrefhttps://ayushi-3536.github.io/deadtreesaerial.github.io/project page. An interactive visualization of the benchmark dataset is also available at hrefhttps://deadtrees.earth/releases/dte-aerial-benchour release page.