discussion / AI for Conservation  / 17 March 2020

Webinar 11PST 3/20 - Deep Learning for Airborne Tree Detection

Hi all, I am part of a regular meetup group on Deep Learning for Environmental Remote Sensing that I think would be of interest to a wide audience. All are welcome. I happen to be presenting this week, but past talks may be of interest as well.

Here is the advert from the organizers:

```
Hi everyone,I hope you are all staying healthy and having productive isolations. We’ll have our next discussion of machine learning for remote sensing applications this Friday at 11am PST / 2pm EST / 8pm CET. This week Ben Weinstein will be discussing his work on tree crown detection using weakly supervised deep learning methods. We can look forward to a fun discussion around multi-sensor data, geographic generalization, and semi-supervision. Here is his latest paper on this topic, as well as a python package (https://deepforest.readthedocs.io/) and benchmark dataset. As always, you can see the schedule of future talks here, and let us know if you would like to present (your ongoing work, a recent paper, etc) at a future meeting! Feel free to share the link to join the group with others. 

Webex link: https://umd.webex.com/meet/hkerner

Cheers,
Hannah and Patrick

```

 




Not that I know of, its a pretty informal group. Happy to post/share slides.

DeepForest docs are here.

https://deepforest.readthedocs.io/

 

Welcome to have a look. My experience is that individual trees cannot be distinguished in satellite imagery. The coarest resolution we've had success with is 0.3m. However, the deepforest weights may still useful as a starting location. If there are visible objects in your image that you want to detect, collecting a few hundred training data samples and retraining the model for 2-3 epochs could be useful. See the link for details. Happy to help, submit issues on the github repo is something isn't clear/doesn't work. Everything is in dev.