After many years of development we are finally ready to share BatSpot, an open-access deep learning framework designed to automate the detection and classification of bat vocalisations from acoustic recordings.
Preprint available here: doi.org/10.64898/202...
Posts by
Great news, AAGE V. JENSEN NATURFOND has granted us 2.5 million Danish kroner to finish the development of our monitoring set-up for bats in Mønsted Kalkgruber and do an analysis of the first five years of data. This is a major boost for our ability to monitor bats in Denmark!
Ever wondered if female birds sing? They do! And more than you think. Read more about how well an automated approach can be used to quantify this interesting behaviour: doi.org/10.1163/1568.... I only had a small part in this study, but Sarah and Karan really did an amazing job!
⭐PhD position available!⭐
Come join us @imprs-qbee.bsky.social to study communication and collective behavior in animal groups! We're looking for someone excited to use computational approaches to tackle biological questions, using our full-group tracking datasets
imprs-qbee.mpg.de/121465/analy...
Many years ago, as a research assistant for Dalia A. Conde, I started working on some amazing data from Species360. Now the paper is finally published, thanks to the hard work from Johanna Staerk and many others. If you wondered why women live longer than men read more here: doi.org/10.1126/scia....
Thanks to @senarjc.bsky.social, @lucymaplin.bsky.social, Mary Brooke McElreath, @mpi-animalbehav.bsky.social, Museu de Ciències Naturals de Barcelona, @imprs-qbee.bsky.social and @cbehav.bsky.social and @royalsociety.org!
Our latest paper just came out. We found that more social parrots have a better vocabulary and females produce more variable call types than males. You can read more here: doi.org/10.1098/rsos.... (PS: the bird in the picture is not a parrot but the cover of RSOS.)
Do you want to get experience as reviewer for a great journal, the Editorial Preprint Team might be something for you!
Running a python script on Windows when you are me: install R, install RStudio, install Git, pull your repo, run python script from Terminal in RStudio.
Yes, and how careful you are. In my defense, Stephen also had problems with most of his disks.
5-6 years would be great. I had four broken before I finished my PhD. SSD is the only thing I can say. Really worth it in the end!
Anybody that uses model selection for causal inference please run this:
a = rnorm(1e3, 0, 1)
b = rnorm(1e3, 0, 1)
c = rnorm(1e3, a + b, 1)
fit1 = lm(a ~ b)
fit2 = lm(a ~ b + c)
summary(fit1)
summary(fit2)
AIC(fit1, fit2)
After a long period of just ignoring Twitter, I'm trying out this platform. Great to see so many familiar faces again!
Also, I moved to Denmark, started a post-doc working on bats, and we have now published our second report. This one looks at bat activity above the North Sea! tinyurl.com/2f8wawmw