How do other species affect the classification?

[Svardsten53]

Many of the recordings used as a basis for classification also include calls from other species. How will these impact the risk of false positives?

[tphakala]

How do you know that? Is dataset used to train BirdNET default classifier available somewhere?

[Svardsten53]

I don't know, that's why I asked the question. I assume that the basis for the classification consists of a large number of sound files from xeno-canto and other. If you then make a classification for a certain species and other species are also heard on these recordings, these should also be included in the "soundscape" and thus perhaps give some false identifications if you don't have a system to handle this. In that case, it would be interesting to know how.