Hi Aaron, a bit slow I was to realise it was Carbon Artificial Intelligence, not Carbon Al as in Alistair. As well as the usefulness in CNM, if you have anything to help process thousands of images from camera traps, I'm interested. Even merely separating images with animals from images triggered by moving tree shadows or moving grass, would save vast amounts of time. Identifying the individual species will require training images. I could help with images from wildlife cameras, if you want those. I am assembling a set of images on a Google Drive to help volunteers. My current camera survey for Rosenberg's Goannas on Ainslie-Majura looks like it will generate more than 1.5 M images. Even though ~30 volunteers are helping maintain the cameras and some are also helping process images, its too much work. In a week, some cameras have generated 15,000 images of moving grass, which is not fun to look through. So the sooner we can do it using AI the better. cheers
That is huge @DonFletcher - top work by the way, but that's an epic amount of data to process for sure.
Agree your with sentiments, we want to use the "AI" as a trained grunt worker, to lighten the load of the skilled expert human moderators who's time is very limited and precious.
We'll be continually training CarbonAI over the coming weeks/months and hope to get it really humming along so it can be super useful for the NatureMapr community.
Describe how you intend to use these images and/or audio files and your request will be sent to the author for consideration.
Your request has been successfully submitted to the author for consideration.
1,902,856 sightings of 21,201 species in 9,373 locations from 13,041 contributors CCA 3.0 | privacy
We acknowledge the Traditional Owners of this land and acknowledge their continuing connection to their culture. We pay our respects to their Elders past and present.