What Wildbook should this feature be in?
GiraffeSpotter
What would you like to see?
We would want to upload data collected from trail cameras to run through the individual identification process. Since the cameras are stationary, some problems have arisen in the past due to the background vegetation matching rather than the giraffe in the images. Additionally, we do have many images with multiple individuals that has been a problem for the matching process in the past.
How would this functionality help you?
This could allow us to expand the database through use of this remotely collected data. We have a large subset of camera trap data that has not been able to be used as of now.
Hey @jenna ,
Thanks for posting. I want to clarify: you’re requesting advancement of the machine learning with regard to camera trap data? Or are you interested in a separate workflow aimed at camera trap data?
Just wanted to check in to see if further clarification on what I was hoping for would be helpful in advancing this?
Currently, in GiraffeSpotter, the workflow has had problems with camera trap data due to matching background vegetation or having trouble parsing out images with multiple individuals. So we were just hoping to provide some camera trap data to help train the algorithm so we can upload more camera trap data to the database.