Fin-drr: Connecting Solemates through Science
There’s an element of serendipity when it comes to audacious ideas. Things come together in a way that seems ridiculous at first but looks like fate in hindsight.
Such were the origins of Fin-drr.
pHacktory meetings are often a mix of random ideas loosely connected. Most get rejected, some hang on for further investigation.
So when a throwaway comment about what a dating app for fish would be called — Fin-drr of course — reminded Andrew Pelling of a University of Ottawa researcher he knew who was tracking fish mating habits, it was love at first sight.
Enter Vance Trudeau, University of Ottawa Research Chair in Neuroendocrinology, and Kimberly Mitchell, PhD Candidate, who were studying the way Zebra fish move when they find that special someone.
Trudeau and Mitchell had hours of video of the movement patterns of these fish, and had previously been doing much of the tracking manually. A perfect place for a little audacious thinking. What if we could tell when a zebrafish had “swiped right” on a mate just by their movements?
With the help of programmer, Eric Jones, the Fin-drr team is embarking on a journey to see if we can predict the perfect match by tracing only movement patterns.
Could this kind of technology be applied to other animals? Perhaps. Could you one day leave the nightclub and be algorithmically paired up with your perfect mate based on your movements throughout the evening? Sounds a little out there.
Over the next 100 days the Fin-drr team is going to explore this fusion of data, computer vision, motion tracking technology and the science of fish mating.
We will be tracking the Fin-drr team’s progress through the next few months. Here are some milestones we are expecting to hit:
- Next week: official Fin-drr lab tour at the University of Ottawa with team members.
- First month: completion of prototype code to be tested on data sets. Work to begin on quantifying motion with tracking technology.
- Second month: test data to be validated alongside human-derived data (hand counted). Work will begin on new zebrafish datasets.
- Third month: creation of a single program for complete analysis of fish movements to be uploaded to GitHub. Start blind tests with fish under less controlled conditions to validate.
- 100 days: produce final report which will include key motions patterns and be released to the open source community for further study.
We are looking forward to what the next 100 days have in store. This project might flounder, it might find its porpoise, but one thing is certain, we’re going to have a lot of fun trying.
