We have a project to build a taxonomy of canid howling behaviour: wolves, coyotes, dingos, domestic dogs, etc. As we have so much data, and it's so hard to analyse these spectrograms by hand, we've set up a crowdsourcing website (http://howlcoder.appspot.com) that allows anyone to help with our project. Users are presented with a spectrogram, and a playback of three seconds of howl, and are asked to draw the frequency profile on the spectrogram image. This way, we leverage the power of human visual interpretation, where automated algorithms just aren't good enough to succeed. We're hoping that this will allow us to characterise the howls of these different species, and draw conclusions about inter- and intra-specific differences.
Please have a look at the site, and share it with colleagues, students, etc. We will be presenting preliminary results at the International Society for Behavioural Ecology meeting in New York in July. The address of the site is: http://howlcoder.appspot.com Thanks, Arik Kershenbaum ============== Dr Arik Kershenbaum National Institute for Mathematical and Biological Synthesis University of Tennessee http://www.nimbios.org/~arik
