Hi everyone! 

We've now posted Sara's session to our youtube channel, and I've also popped it up the top of this thread.

The collaborative notes worked really well! I've now updated them to capture what happened in the chat - it should be a helpful companion to go alongside the recording. The notes have links, projects, and key discussions we saw in the chat, and summarise the questions Sara coverd in the discussion as well as the Qs we weren't able to get to (40mins overtime was our limit!). If your question was one of the outstanding ones and you'd like to have it answered, please drop it in the discussion below. 

The notes now also have the participant check ins (such an awesome range of places, projects and interests!) - I'm sharing these as seeing what other people are doing might help you connect with each other. If you see someone you want to connect with, try and find them using  our member direcyour people tab. If you can't, email Ellie and she will see if that person is happy to hear from you before connecting you.

Reminder, registration is open for Carlos' tutorial next week: How do I perform automated recordings of bird assemblages? Register here.  

Thanks everyone! 

Steph 

Great talk! I thoroughly enjoyed it. Some  high schoolers have done small AI projects(s) and have interest in the wildlife.
What resources would you all suggest to further develop high schooler’s interest in AI?
 

Hi everyone,

I am resurrecting this thread as WILDLABS is currently planning an online tutorial on camera trap repairs in the field on 23rd July at 11 AM ET on which I'll be co-presenting. Do join then if the topic is still of interest!  Also, if you've got repair tips you'd be willing to share on this webinar, possibly as a short step by step video, please get in touch.

Diagnosing the cause of malfunctions is a recurring theme on this thread but are there other issues you would like to learn more about or malfunctions you encounter regularly? I created this spreadsheet a while back to get an idea of what the most common camera models and failures were. if you are interested and have a couple minutes, could you contribute to it? It would really help make the tutorial webinar as focused and relevant as possible and maybe create more tutorials and ressources at a later stage.

Hi Michelle,

I had a group of undergrads help me with a 40,000-image dataset a few years back. We used the TEAM network Wild.ID program, so each photo that was tagged indicated who tagged it. That was helpful for checking quality later on. For our common, unmistakeable species (e.g. whitetail deer), I didn't require a second identification, but for more challenging groups (foxes, mustelids), I would often have a second person review the ID, or do it myself. Later on, I had a student go through all the tagged images of a particular species (gray squirrel, etc.) and verify the first ID. I found that some of the undergrads were very reliable in their ability to ID the species, whereas some other students needed to have their work checked more meticulously. I later thought of the idea of building a training set of say, 100 photos, to have each student run through to get a sense for their familiarity with the species, but also their ability to handle the more tricky scenarios that come up often in camtrap datasets.

Most folks could only handle 1-1.5 hours of continuous tagging. I had a few enthusiasts who would go for 2 hours straight, but that was rare. We logged effort in a shared google spreadsheet, where the students noted the dataset they worked on, any issues that came up, and any individual images that needed a second check.

I also tried to set up a more ergonomic workstation for folks (multiple monitors raised up, ergonomic mouse, etc.). Since the motion is so repetetive, easy for folks to develop carpal tunnel syndrome.

If you are dealing with a much larger dataset, you might want to look into more sophisticated AI/automation methods, but for a smaller project, this was doable. If you have a university connection, you can often recruit folks through chapter groups of The Wildlife Society. Student are often eager to gain experience, although many don't stick with it once they find out how unglamorous it is!

Good luck!

-Andy

Hi Gregory,

That is quite shocking to hear that a road could be so detrimental to wildlife. If you were able to place a speed hump or rumble strip where you place the camera you could even embed the metal detecting sensor in them to ensure accurate detection of every vehicle when they are going slower.

Our camera takes a burst of images and this can be set to be quite a quick burst or a slower burst so with a little bit of testing I am sure we could get the numberplates.

I am assuming that a lot of time these collisions happen at night. Would there be a way to have a fixed light at the camera locations as otherwise the cameras can be blinded by headlights and taillights making reading the VRN difficult.

At the moment our system is satellite connected. This does have a relatively high monthly data cost. We are also planning to develop a cellular version but this will require more funding and development time that we have not got yet.

The system is in the final stages of development and we plan to launch it in mid-2020 once it has undergone some prolonged field trials at ZSL sites, so it is not currently available. Final costings are still under review and will be released early next year.

I am sorry that we are not able to provide this capability to you right now. 

Have you found any commercially available camera systems? In the UK I believe the automatic number plate recognition cameras placed on gantrys over motorways are called SPECS and a quick google said they are sold by a company called Jenoptik. If you find out how much they are and how they work, perhaps you could update this thread?

Thanks,

Sam