Hi I am also looking for collaborators

Read "Looking for Collaborators: AI-Powered Backyard Bee Monitor for Citizen Science | WILDLABS" on

Looking for Collaborators: AI-Powered Backyard Bee Monitor for Citizen Science | WILDLABS

if interested in the project let me know

we can add acoustic monotoring for bees

Hi Leonardo,

Let’s chat! I potentially have some opportunities, collaborations or even short-term employment. Send me a message. 

Hey Leonardo

I might have something. Lets chat, reach out to me at [email protected]

This looks amazing and I look forward to trying it out when I get the chance! 

Just wondering, when it comes to the AI recognitions, is there a way to "rename" the recognitions that were incorrect? 

Hi Kristof,

 Very cool that you made your own app like this! I'm not a movement ecologist, but I'd generally say this sort of thing will be more nice to look at than solves research questions. For example, most statistical models will show that proximity to water and human development are the strongest predictors of elephant movement. But it's still a very cool tool for outreach and communication with the general public or conservation decision makers! 

I can imagine a visualization like this being very helpful in communicating drivers of human-elephant conflict in SE Asia, where elephant movements outside of protected areas may be explained by the combination of attractive food crops, accessible water and shade refuge in tree plantations, and repulsive hazing by farmers. These sorts of things that can be modeled statistically don't necessarily translate well to the public - so visualization is really important! 

On a non-geospatial note, I would say that AR/VR also has a lot of potential for increasing public engagement/support for conservation. Getting individuals who may never go to East Africa to experience visually how increasingly severe droughts affects the landscape and promotes conflict between pastoralists and agriculturalists and wildlife could be really powerful. That's perhaps an area where game development expertise would be particularly useful. 

Cheers,

Brandon

I am glad to see more programmers coming into the conservation field.  The first big project I did that really got me involved with conservation work, was taking the path finding algorithms I used from learning game programming, and using them to detect and measure the distance of routes that turtles traveled up and down streams in a river drainage.  

Wolves, cool!

Will this then need collared wolves ?

Hi Daniel

It depends on what you want to be
If you really want to bring something new then I guess do a PhD

I have a weird background, applied mathematician, PhD in econ and now working on this
I am guessing that the transition is doable as the math is the same that's how I kept doing very different subjects

I am currently working in an french engineering school on international shipping and would love to translate AI for conservation of ocean and I have several ideas

So I would be happy to discuss them and if interested let me know?

Hi @djscrazyhorse 

I've been looking at a similar thing. I actually decided against it in the end, and instead to take a tack of developing a solution for a particular niche.

I just published an article about the thought process here:

https://medium.com/mdes-environmental-social-impact/photo-monitoring-for-the-flax-roots-557749cdb6f3

Would love to hear your thoughts.

Any reason you went with those particular App Builder solutions, out of interest?

If you still need help with any of your backend on this with data.  I could possibly help.  Let me know.

It would be cool to open-source/MIT license a simple utility written in QtXR, (especially for the SMART crowd, since their mobile app is written in Qt..), that could capture and encode video+audio+camera_metadata+6DOF+Video+GPS+SMPTE timecodes to a logfile of an emerging standard off an Android tablet .. for "I am looking at this  thing (description of thing here...) " video to a "well-known" BLOB type in the SMART configurable data model...

and totally awesome to translate the standard file to KML tours , AR tours with Eclipsa streams, playback against NiaticLabs reconstructions, Google XR tours, etc ... 

(extra points ... 😎... for a pipeline that transcribes the audio trough speech->text,  and the video through multi-modal scene processing for species and other metadata ("#cuteanteaters"), and exports it to a digital-rights framework so the authors can get Tik-Tok revenues ... )

... as a no-cost/thriving alternative to what WildLife Insights, GoPro, and others will inevitably do, that locks sites into their subscription models .... 

Hi Jeremy,

I am copying here a reply from Juliane Röder, who is not on the forum here:

The GFBio VAT tool seems to do all the desired analysing and visualising stuff - but it isn't specialized on camera trap data. Instead, it harvests GBIF, GFBio data centers, and environmental data that you can combine in any way you want: https://vat.gfbio.org/#/

6:01

Documentation with some example code and notebooks is here: https://docs.vat.gfbio.org

6:04

A standalone tool for camarea trap data and images is nice, but maybe joining forces and adding some functionality to the VAT tool to better handle camera trap data would be a more sustainable solution...?

The VAT tool can give you some simple statistics of the chosen data (e.g. occurences of wild cats in central Europe in winters 2000-2015), and some simple analyses (wild cat occurences and snow depth). It's focus is on geospatial analyses of occurrence data. They do not have any functionality specific for camera trap data - yet.

My point was, that if you want geospatial analyses and/or additional data to compare to camera trap data, you probably don't have to build a new tool.

I'm not involved with the development of the VAT tool, but I could help with the contact. I know the VAT team.

You can contact her on [email protected].

I did a quick exploration of the VAT tool. It is built quite differently, and my first impression is that it is less accessible; however, as you become more familiar with it, you discover a wide range of options to explore for DwcA occurrences.

I believe it is extremely valuable to also have a tool that works directly with the richest data (Camtrap DP), rather than relying on translated data (Darwin Core), where important details and data structure are lost.

I also think the goals of both tools are different: map-based analysis vs. exploration.

But still usefull to get to know about each other :-)

cool

Thanks for sharing this Margaux! Very interesting to read.

I'm curious whether you've experimented with Claude code CLI or ChatGPT Codex to help with development and how it compares to the Github Copilot plan? At least in my experience, the release of Opus 4.5 in Claude code in late November has felt like another step change in terms of capabilities for these models 

Great presentation, Margaux!

It depends by your team' needs, level of automation you want to achieve, and budget availability.

More expensive but with extra benefits than Copilot is Cursor. However, Cline has the best of both worlds: price and capabilities.

Have you looked into Cline and Bedrock is your stack is AWS?