Interesting site,but question. Your ONLY stated purpose is to study Chimpanzee. For me I am getting about 2% of the videos with chimps. I understand the need to reduce the 7,000 hours down to 140hrs plus of good chimp data, that is our job. Why are we also IDing other things that are not chimps? It would be faster to say "Chimps yes", with info or "Chimps no" and move on (unless rare animal comment). Also, your animal behavior comments are really only for chimps. Also, with 2% hit rate it is hard to develop a personnal good skill set on chimps versus if we studied them after the reduction to chimp videos only.
Yes, our primary goal is to understand chimpanzees, which is why we ask for so much detail from the chimp videos. But it is also important to know the chimpanzee predator density at each site (leopards), the density of certain prey animals (monkeys), the human pressure at each site and the density of species that are food competitors with chimpanzees (elephants, pigs, monkeys and more). These are all factors that directly affect chimpanzee behaviour. Furthermore, we want biodiversity estimates for each site which is why we want each animal identified in broad categories. From this we can also ascertain which species make good indicator species for these forests which help us develop better biomonitoring tools. It would also be a shame to waste this opportunity to know what is in each video just because we may not be directly interested in it. People who study duikers will be super interested in this data for example and once all the duiker videos are identified we can easily pass these videos on to them for their analyses.
I think if your main interest is understanding chimps and telling them apart, heading to the chimp ID boards is the place for you. This is basically where all the chimp videos are reduced as you describe. It is tough telling these guys apart from camera trap videos and the more dedicated help we have with this very specific task the better!
Will any of the classification results be publicly available once all videos are sufficiently classified, or will it be kept private (and distributed only to trusted parties) to reduce likelihood of poaching, since I imagine results would include geocodes?
That's a very good question. Typically, we open-source all data once a project is completed. But this project may merit particular caution in that regard. We'll probably address that a little further down the road, but we definitely will keep poaching concerns in mind.