DeepTrees Symposium: Deep Learning and Remote Sensing in Tree Monitoring

Monday 2 Dec 2024, 09:00 → 16:00 Europe/Berlin

Halle 1 CD (KUBUS)

How can DeepTrees improve biodiversity monitoring?

What role does machine learning play in automating urban forestry management?

How does DeepTrees contribute global conservation efforts?

 

The DeepTrees Symposium will focus on cutting-edge advancements in machine learning and computer vision for tree monitoring, particularly leveraging deep learning models for automated tree inventorying from orthoimages, focusing on the Digital Orthophotos (DOP) aerial imaging program of various federal states in Germany. Tree monitoring is essential for urban planning, environmental conservation, and forestry management. The symposium will highlight the project’s goals, including standardized workflows for tree crown detection, analysis of tree allometry and classes, and the creation of a database to support local environmental monitoring efforts in the Federal States of Saxony and Saxony-Anhalt in Germany. The symposium will bring together experts from various fields to discuss DeepTrees updates, community results, and future use cases for the system in biodiversity assessment, urban forest management, and ecosystem services modeling.

flyer.pdf

Website

09:00 → 09:30 Opening Remarks + Keynote

Keynote speaker: Hannes Feilhauer
Title: Tba

09:30 → 10:15 Deeptrees updates

This session will focus on recent advancements in DeepTrees, including updates on workflows, tools, and methodology improvements. Talks that highlight:

• Innovations in deep learning models for tree detection and monitoring.
• Workflow standardization and methods for aggregating tree data.
• New findings in tree allometry and tree class analysis using DeepTrees tools.

10:15 → 10:45 Break

10:45 → 12:15 Results from the community

This session will highlight community-driven results, with a particular focus on how researchers, city planners, and environmentalists are their own work. We encourage submissions for:

• Case studies in urban forestry, biodiversity, or ecosystem management.
• Results of studies involving remote sesing and or machine learning for tree/forest monitoring.
• Comparative analyses of different deep learning models (e.g., CNN, UNet) used for tree monitoring.

12:15 → 12:30 Group Discussion Buffer

12:30 → 13:30 Lunch (Mensa, not catered)

13:30 → 14:30 Potential use cases

This session invites forward-looking presentations discussing the real-world applications of DeepTrees. Potential speakers are encouraged to submit sessions on:

• Enhancing urban forestry management and ecosystem service modelling using tree crown data.
• Potential of DeepTrees for improving biodiversity assessments and carbon storage estimations.
• Innovative and novel use cases for DeepTrees in environmental monitoring and conservation projects.

14:30 → 15:00 Group Discussion Buffer

15:00 → 16:00 Workshop: Hands-on with DeepTrees Workflows

Interactive session on using DeepTrees python package for tree detection and analysis. Attendees will engage with real data to explore tree monitoring processes using our code library.