As each day goes by, real life and technology become more unified than ever before. And in those areas where they aren’t, it feels kinda weird. For example, did you ever drive by a billboard or some notable sight on a highway but your passenger missed it? For a split second, do you think, “Oh, just rewind it for her,” as if real life were a DVR?
I get the same feeling when I hear birds chirping. “What species of bird is that,” I often wonder. Followed by, “Why isn’t there a Shazam for bird calls?” Why can’t I launch an app, hold my cellphone up, and have it tell me if I am hearing a warbler or a sparrow, a thrush or a grosbeak?
Turns out, scientists are working on it, but it’s taking more time than anticipated for a variety of reasons. We spoke to one of the preeminent researchers in this area, Dr. Mark Berres, assistant professor of avian biology with the Department of Animal Sciences at The University of Wisconsin-Madison, about the Shazam-like app he’s been developing for a few years, called WeBIRD.
Technology Tell: What do you hope will result if this project comes to fruition? In an ideal world, what would this app accomplish beyond its core function of matching field recordings with a bird call database?
Dr. Mark Berres: I hope that it will become a learning tool that will foster an enhanced connection between individuals and their environment. Many environmental issues our society faces are rooted in unfamiliarity with it. Adults and children are strongly influenced by the media, a phenomenon known to extend into attitudes concerning wildlife. Logically, if proliferation of electronic media continues and environmental education is deemed essential to attenuate the continuing trend of biodiversity loss, we must consider the capabilities of specific media and use them appropriately. Failure to do so may have unintended consequences early in life, all of which are difficult to reverse. For example, most information acquired by children about animals is now obtained from websites. Research has demonstrated that exposure to this media created strong biases where considerations to protect a species were driven only by a few iconic and exotic species: the diversity of species that should be protected was small and not local.
TT: Where does the WeBIRD project stand as of now?
Berres: At this point I am focusing on other aspects of WeBIRD’s implementation and use as a learning platform in a limited geographic region (southern Wisconsin). You can take a look at an early version of our efforts by going to the Apple AppStore, search for ARIS (the GLS mobile learning tool), install it and then search for WeBIRD (within ARIS). The “quest” version is much more interesting that the other one (the two versions were designed to test specific learning objectives and how difference between student performance were manifested). Note that the ID tool is not implemented in these versions.
TT: What kind of feedback have you gotten about the WeBIRD project?
Berres: We use it in our research to monitor birds in the field in near real-time (this is what it was originally developed for). We have also began to integrate it into one of my courses “Birds of Southern Wisconsin” as a learning tool. Although there has been no official public release, the amount of interest from the public continues to be substantial. Unfortunately, I do not yet have anything to give them because of a peculiar, but not unexpected, problem. Nevertheless, I am preparing a website where people will be able to learn more about WeBIRD. Among other fundamental audio analysis information, users will have an opportunity to submit their own recordings of avian vocalizations and test the identification procedure using the current audio database (southern Wisconsin bias implied…).
TT: Are there currently any other/competing apps or projects in development that you’re aware of? If so, how does WeBIRD compare?
Berres: Not really, at least at the intent and scale of what I am trying to accomplish. The article I emailed you mentioned one “extinct” device (SongSleuth) and another rudimentary app that I have not yet examined.
TT: Shazam can recognize hundreds of thousands of unique songs. Why is it so difficult to develop an app that recognizes thousands of unique bird calls?
Berres: This is the million dollar question with an easy answer: Shazam has a database of originals to which it compares against queries that are nearly identical and high-quality copies of the originals. This cannot occur in natural vocalizations as no “original” exists and there is extensive variation among individuals and further structured by geographic region (much like human accents and dialects). So while WeBIRD can ID birds quite well, it requires a database populated with exemplars of vocalizations. Getting these vocalizations is very HARD (which is why I am attempting to crowd-source these vocalizations).
TT: I love the idea of crowdsourcing data for the app. Can you elaborate on that angle?
Berres: Since vocalization data is crowd-sourced, WeBIRD will provide scientists with a return of high-quality data that no team of professional researchers could ever hope to achieve. As more audio data is collected, the reliability and scientific value of WeBIRD increases. Moreover, because WeBIRD collects and stores the actual audio data used to determine the species, the scientific value of citizen-collected data is on par with that collected by professionals and eliminates a detracting reliability concern.
The capabilities of WeBIRD will provide scientists with a new tool to integrate into their research. Tasks ranging from remote field monitoring to fine-scale analysis of vocalizations will become easier and more productive.
Generally speaking, media theorists submit that new media – specifically mobile media – affords educational and research experiences that in many cases were previously unobtainable. In the case of WeBIRD, the novelty lies in the integration of automatic birdsong recognition with crowd-sourced, location-aware mobile devices. For informal users, learning the species names of birds (and other organisms) is a necessary yet difficult first step to cultivate a positive relationship with nature and more broadly, the environment. By providing immediate feedback while actively engaged in birdwatching, WeBIRD will help anyone using it to identify birds accurately and improve their identification skills. An improved ability to identify and monitor birds, such as I promote here, will enhance scientific research and ultimately afford greater awareness of the public to the biodiversity concerns both at local and global scales.