Audio clip submissions for detection model improvement?

[tphakala]

Do you accept audio clip submissions for model improvement?
For example following dog bark is detected at 99% accuracy as Eurasian eagle-owl, in fact 90% of owl detections are dogs in my samples.

bubo_bubo_99p_20240203T215153Z.zip

[pddpauw]

Yes, would like to know indeed... I recently had several Grus grus identifications (up to 98% confidence level), but our birder community is sure this should be Anser anser.
Where could we submit the files for this?

[tphakala]

Is there any hope we could help improving prediction patterns by providing incorrectly classified audio files?

[tphakala]

I have solved this by training my own classifier to improve dog etc. detection