Bird Recognition in the City of Peacetopia (Case Study)

Question about what to do with the added 1000 images of a new species is somewhat incorrect. Ir is at least up for debate.

I know the it’s about a made up place Peacetopia which could or could not be a place on Earth. Doesn’t say but since I’m from Earth I have to use my point of references :slight_smile:

If you look at the number of bird species (on Earth) 18.000 from lastest studies.


Given the 10.000.000 dataset I came to the conclusion there was about
10.000.000 / 18.000 = ~ 555 images per species if the sampled uniformly.

So adding a 1000 new images would be enough to warrant it’s existence in the training data.

I think the question is written by someone biased (not having enough interest in birds :-)). :slight_smile:

Just my thinking


This question made my day and I definitely cannot think of any good answer :grinning_face_with_smiling_eyes:

Anybody else of the moderators/mentors want to chime in with some input?

To discuss further, we need to know how many bird species may be found at the same place on Earth. A city like Peacetopia :smiley:

I guess that the monitoring system wont be a world market product but rather a one-off. Would be nice for the people of Peacetopia to be able to recoup some of the development costs by selling it worlwide. Maybe even make a profit and offer free healthcare and take proper care of the elders :slight_smile: