Working Groups 2 and 5
Data preparation for Deep learning models
Duration: 3 months.
Timing: 20 January 2025 (flexible starting).
Location: in person
Total estimated grant: 4000 euros
Description: This STSM Position Grant is important to the WG2 (data fusion) and WP5 (forest ecology) activities. During this STSM the successful candidate will help with the website preparation, data cleaning, segmentation of the trees, and tree species annotations for deep learning models. The project focuses on mapping individual tree species which is a critical task for forestry management. Using aerial RGB imagery, we aim to create a cost-effective, automated system for detecting and identifying tree species, with broad applications in forest monitoring.
Specific tasks include checking the alignment of tree positions with ortho-mosaic, checking the species’ Latin name, delineating and annotating tree crowns, etc. The database is built based on this data call: https://3dforecotech.eu/activities/tree-species-annotations-and-aerial-rgb-or-rgbi-imagery/
The candidate is expected to have up to 3 months of 60-80% time availability for this project. Experience in using GIS software e.g. ArcGIS Pro or Python and working with shapefile and raster data is required. Knowledge of segmentation approaches is needed.
Anticipated output: The final output will be tree canopy delineations. These will be used further to train a deep-learning model.
If you are interested, please send your CV and a short motivation with an example of your work with segmentation methods for raster images to the following e-mail addresses:
- Zhongyu Xia | zhongyu.xia@usys.ethz.ch
- Mirela Beloiu Schwenke | mirela.beloiu@usys.ethz.ch
- 3DForEcoTech | 3dforecotech@gmail.com
Application deadline: 06.01.2025. Applications received by 06 January 2025 will receive full consideration, however, the position will remain open till filled.
