Minnesota Deer vs. African Antelope

Written by Emma Bublitz, the social media intern for Eyes on the Wild and social media manager for Snapshot Safari. 

The picture on the left is a buck from the Eyes on the Wild project. The picture on the right is of a kudu in Mountain Zebra National Park from Snapshot Safari.

There are a lot of great camera trap projects, and many of them ask citizen scientists for help in classifying images. Researchers from the University of Minnesota are proud to offer several projects, including Eyes on the Wild and Snapshot Safari. I feel incredibly lucky to be a part of both and can say that I like them both for different (and some of the same) reasons.

As some of you may know, Cedar Creek Ecosystem Science Reserve (where the Eyes on the Wild project takes place) is a field station for the University of Minnesota College of Biological Sciences, located in East Bethel, MN (~45 miles outside of the Twin Cities). Researchers at Cedar Creek has been trying to understand ecosystems and their functions for decades, but their attention has recently switched to the wildlife on the reserve. This is where Eyes on the Wild comes in. Documenting animal sightings from the camera trap images will give the researchers valuable information about the abundance and behavior of the animals at Cedar Creek.

A red fox strolling through Cedar Creek.

What you may not know is that another University of Minnesota research team, the UMN Lion Center, has released two highly successful citizen science camera trap projects. Snapshot Serengeti was created in 2010 and utilized camera traps in Serengeti National Park in Tanzania. Snapshot Safari was released just over a year ago and features camera trap pictures from the Serengeti, along with reserves and national parks in Mozambique, South Africa, Botswana, and Zimbabwe. Snapshot Safari aims to monitor the health of critical wildlife areas in Africa in an effort to learn how to best conserve the species who call these areas home. 

A lion selfie from Pilanesberg National Park.

I've spent a lot of time on both projects and can say that I thoroughly enjoy both of them. I love that they both offer you a peek into the lives of wildlife, showing volunteers animal behaviors and personalities that they might not ever be lucky enough to see in person. I have also seen first-hand the tremendous impact that the classifications from both projects have in scientific research and education.

Even though the interface and reason for creating both projects are similar, the projects themselves vary tremendously. Snapshot Safari's camera trap images from across sub-Saharan Africa feature antelope species, big cats, zebra, giraffe, ostrich, baboons, and wildebeest, among others. It is absolutely incredible that citizen scientists can see animals they'd see on a safari from their computer screen.

Another perk of Snapshot Safari is that there is a variation of site options, each with their own species and landscape. For example, Snapshot Safari currently has images from the Mountain Zebra National Park, Pilanesberg National Park, Grumeti, Serengeti National Park, and Ruaha Carnivore project available. There were also images from other sites, but they have since been retired or the team is working on uploading more. Some Eyes on the Wild users may get tired of seeing the same road, tree, or field, but being able to choose which Snapshot Safari site you want to classify offers users the opportunity to switch things up.

A baboon strolling through one of the Snapshot Safari sites.

With the wow factor comes more, sometimes very difficult species to identify. The classifying menu has 66 options to choose from, while Eyes on the Wild has only 30. Species seen on cameras are often less well-known than those in Eyes on the Wild images. However, learning to identify these African species also provides a unique opportunity to learn more about the ecosystems. I truly have learned so much by being a part of this project.

Eyes on the Wild images are much different than those of Snapshot Safari. You won't find lions or giraffes, but instead, you will find deer, raccoons, insects, black bears, squirrels, several different types of birds, and occasional canids, among others. Although these may not be as exciting to see for some as the African wildlife, I personally love the local aspect of Eyes on the Wild. I'm from a suburb of Milwaukee, WI and grew up seeing and being fascinated by a lot of these animals. However, what I didn't get to see growing up was how these animals allocate their time throughout the day. I saw deer munching on the side of the road, but never knew that they spent their winters nibbling on cedar trips. I saw squirrels but wondered how they survived the winter. Eyes on the Wild images answers these questions. It has also taught me to think and want to learn more about the lives of the animals that I see and encounter on a regular basis.

A black bear selfie from Eyes on the Wild.

One unfortunate similarity is that both projects have the dreaded "Nothing there" or "No animals present" images, and there isn't currently a way to get rid of those completely. However, I can assure you that both project teams are doing their absolute best to try and minimize these annoying pictures! Besides trying to ensure that vegetation does not grow in front of the cameras, both the Snapshot Safari and the Eyes on the Wild teams have been working on incorporating machine learning into their interface. Much of that work is being done in tandem, with the same computer scientists contributing to both projects! I'm definitely not an expert, but it is my understanding that once the computer has been trained, it will be able to identify whether there are any animals present in a picture. If the computer doesn't detect any animals in the picture, the image will never get shown to volunteers. This will likely be a long process, however, since training the computer can be difficult. In the meantime, we should soon have the ability for the computer to make a tentative assignment - of a person, a car, a specific animal, or 'nothing there'. For non-human non-animal captures, a single volunteer classification that matches what the computer had assigned will be enough to retire an image. By that method, even though the grass pictures are still in the workflow, they will get retired much more quickly and soon our volunteers will be able to just focus their energy on wildlife! You can read about some of this in-progress work in a synopsis of a paper UMN scientists published just last month. The bottom line is that scientists on both projects agree with you that those 'nothing there' photos are a pain, and are doing their best to maximize users' valuable time spent on the project!

Eyes on the Wild and Snapshot Safari are both incredible projects that offer people the opportunity to learn more about wildlife. Neither of them would be possible without the support of our enormous, global volunteer base. Keep up the great work! If you have any questions about either project, feel free to email me at bubli018@umn.edu. I'd be happy to give them my best shot!