Saul Greenberg is a great guy! He has made a few very useful videos regarding the Timelapse/Megadetector integration:

https://grouplab.cpsc.ucalgary.ca/grouplab/uploads/Publications/Publications/2021-05-ImageRecognition-Video.mp4

and a tutorial: 

https://saul.cpsc.ucalgary.ca/timelapse/uploads/Videos/Video-TimelapseImageRecognitionGuide.mp4

@bluevalhalla There will be a amazon classifier (PyTorch Wildlife) added to the list soon (hopefully this week). 

Adding the animl-py models is on my TODO list, but unfortunately the TODO list is quite long....

At this point there is no option of visualising video's. What you are seeing is the direct result of MD on videos. It would be possible if we keep track of which bounding box corresponds to which frame. Perhaps a feature which exports the frame with the highest confidence per video would be sufficient. But then again, at this point I don't have a lot of time to work on that. 

Same with CamtrapDP, it's on my TODO list!

Thanks for your insights Kim!

But the use of the browser password manager should make it no big deal for people even if they got logged out once a day. Which I’m sure is not the case for the wildlabs site.'

"Big deal" is very subjective...

I use firefox on linux and have tabs unloaded when not in use for ~30 minutes. After 24-48 hrs if I click on the wildlabs tab I'm logged out. I then have to click on login and have my password manager fill out the form. Then it brings me to my profile page (why? why doesn't it go back to where I was which is where I wanted to be, just logged in?), then I have to click on the feed page, which displays the global feed which is also not of my interest. Finally I get to click on my feed. Is is a "big deal"? I can't claim it is. Does it keep me from checking wildlabs? Definitely!  I participate in a dozen forums/discords and wildlabs is the most annoying to check in for a quick "oh, let's see whether there's something interesting".

It's still a choice to close your tabs. Closing all your tabs tends to get you to have to login again.

However.... Despite the expiry date on the cookie. linkedin.com logs in without ever having to re-enter your password. So I guess there is a way in principle to make it so you don't have to log back in. Such would a way then also work with your auto tab closer.

However, that's not something that I'm going to investigate. If someone wanted to, the clue would be in the cookies and the final answer would be available from the wildlabs site developer. Myself, I would never have guessed that so many people would find this annoying. But everyone is different.

Hi Ned, the TTP223 is commonly available in a small module, so I'll use that instead of making my own board just for that chip.  Probably I'll wrap it in PVC heat shrink tubing (often used for batteries) so I can cement it to the inside bottom (flat) of the PVC waterproof housing.  An alternative is to integrate the chip (SOT23) on main board and wire its capacitive input to a patch of adhesive copper foil on the inside of the hull, but that's getting a bit too fancy for me.

I'll configure the module output active high and connect it to the MCU's wkup pin (also active high) with a pigtail.

The plan is to allow the MCU to boot up and do the self test before putting itself to sleep, all while on the bench and before the housing is closed, and to wake up upon deployment.  I may have the code poll the touch sensor during normal operation to detect recovery in order to put it back to sleep, or I may not bother.

Actually my Raspberry Pi application is a sound localizer not related to image recognition. My image recognition related project runs on Jetsons and higher.

But I think recognizing bugs on a drone would likely be challenging. You would have to have sufficient detail to get good recognition which would be a very narrow field of view and then vibration also becomes an issue.

For example, the trainings on just the coco dataset seems to distill the recognition of people to a multi-segmented thing with bits sticking out. So spiders on camera lens are highly likely to be seen as people. To get better results much more training data is needed. I expect it's also likely to be the case for insects, really large amounts of training data would be needed to tell the difference between different types.

Dear Akiba.

Sorry for late reply.

I'll contact you via DM.

Hi Cathy! Please check out the below article on group managers. Essentially, we're asking for a 12 month commitment of 1 hour a week (more or less) to promote engagement in WILDLABS groups by sparking conversations, engaging with people's content, and planning 1 virtual event a month to bring the community together. (This could be bringing a speaker in, or just having a 45 minute coffee call for people in the sector to get to know each other and share their work.)

As a group manager, you have the full support of the WILDLABS community team. We have a dedicated Slack channel, monthly calls to support you in your group management, and so much more.

Being a group manager is a great way to give back to the conservation tech community, steer conversations in the direction you think is important, foster a vibrant space for collaboration, and build your network. Shoot me an email if you're interested ([email protected])! I'm happy to hop on a call to discuss more in-depth. 

Unfortunately I may fail to deliver so I cannot take it up. Am off net at times for two weeks straight. 

But excited about everything GIS so I follow keenly. 

Thank you for the explanation. 🤍

Hi Emma,

Thank you for your outline. Feel free to send me an outline ( I sent you my email in a private WildLabs message ), but from your description I would like to give two comments already,

1) Besides learning about these technologies, the course should in my view also include learning the skills of finding out about new technologies and assessing them. The list of technologies that you are giving seems state of the art, but in the course of the students career it is likely to change completely a couple of times over. Much of what is on the list would not be there ten years ago. Even only three years ago AI (as we now think of it ) would probably not be on the list, but ML would.

2) Perhaps you were planning this already, but when it comes to 'correct use and validation', in my mind, that should definitely include the social/societal aspects of these technologies, no matter if it is about use in the global North or global South. What I am after is Technology Assessment.

Hopefully, this is of use to you

cheers

Frank

Building on Frank's comments, I'd also think about including some conversation about the ethical use of technology - Trishant's talk at Tech Tutors a little while ago sparked a really important conversation here in our community, one that would be valueable to have with your students.