Our Citizen Scientists have been religious sharing data for the past three months, and helping trouble-shoot the issues with the cameras. A few of cameras have had to be brought back to the lab following heavy rains but most of them are still standing strong and collecting precious footage.
At Deakin we are getting ready to provide a report on what species were found on during the first month, personalized for each participants. We are a tiny bit late because of the latest snap lockdown that prevented access to the campus for a few weeks.
Besides the first month, the other months will also be analyzed in the future… but the algorithm is still a work in progress at this stage. Each video is analyzed frame by frame and that takes a fair bit of time, even when the computer is doing the job! At this stage, the results also need to be checked ‘manually’ to make sure the algorithm is performing well.
To do so, the videos analyzed will be set up on a Citizen Science platform (Zooniverse) for the participants to help evaluate how well the algorithm is performing so far : do we still have a lot of videos without animals in them? Are the labels corresponding to the animal(s) in the video?
That will also be a nice opportunity to see what other citizen scientists have been finding on their land!
Above : some pictures from a commercial camera, that we use as a ‘control’. That is why you can see our TechnEcology camera in each frame. Those pictures feature a bare-nosed wombat (Vombatus ursinus), swamp wallabies (Wallabia bicolor), a short-beaked echidna (Tachyglossus aculeatus) and a brush-tailed phascogale or tuan (Phascogale tapoatafa). The latter is classified as vulnerable in Victoria.
Leave A Comment