WILDLABS logo

WILDLABS logo

Group

AI for Conservation / Feed

Artificial intelligence is increasingly being used in the field to analyse information collected by wildlife conservationists, from camera trap and satellite images to audio recordings. AI can learn how to identify which photos out of thousands contain rare species; or pinpoint an animal call out of hours of field recordings - hugely reducing the manual labour required to collect vital conservation data.

discussion

Global model for Livestock detection in airborne imagery - Data, Applications, and Needs

benweinstein

12 August 2024 8:19pm

Hi all,I was at a AI for ecology working group a few weeks ago and was asked to look into an airborne model for detecting livestock to assist in land-management, agriculture and...

2 2
Elsa Carla De Grandi

12 August 2024 11:03pm

Hi @benweinstein,

An open-source detector would be great!  from my point of view it would be ideal if it could be applied to Very High Resolution satellite imagery.

Maybe you are aware of this already but there seems to be a group at Missisipi State University who is putting together a platform for Aerial Wildlife Image Repository (AWIR) to promote research and storage of wildlife imagery and data (I haven't looked into it - it seems you'd need to sign up but looks potentially useful).

They have recently published a paper and their dataset is open source.

I would also suggest posting this discussion in the "Remote Sensing & GIS" and the "Drone" group.

Cheers, 

Elsa

benweinstein

14 August 2024 7:10pm

Thanks, I grabbed those datasets already. I know about that website, put couldn't quite make sense of it. It says 'open repository', but then there is a login page with no way to request to join as far as I can tell. 

See full post
discussion

scikit-maad community

Haupert

8 August 2024 10:16am

 We, @jsulloa and @Sylvain_H, are the main contributors of scikit-maad, an open source Python package dedicated to the quantitative analysis of environmental audio recordings...

10 5
Juan Sebastián Ulloa

11 August 2024 8:19pm

We're excited to announce that a new release is now available, scikit-maad v1.4.3! This update includes support for Python 3.11, along with bug fixes. Make sure to update to the latest version to take advantage of these enhancements.

Arky

14 August 2024 6:01am

Hey @Sylvain_H 

 

Am leaning towards Maad Audiophiles or The Maad Eco Community

 

See full post
discussion

How do drones help agriculture?

rainy

25 June 2024 8:26am

How do drones help agriculture?In addition to monitoring crops in real-time, AG drones can also be used for other tasks, such as crop spraying and seeding operations...

4 1
Sarah Mshanga

5 August 2024 6:23pm

 

In addition to what has been mentioned above, also drones help in the early detection of pests and diseases, allowing for timely intervention. They can also plant seeds, reducing labor costs and increasing efficiency.

Steffen Döring

9 August 2024 12:21pm

Hello,
also in Nature Conservation and Biomonitoring drones can help to reduce labor costs and increase efficiency 😉

My free manual 'Drones in Biomonitoring' - https://doi.org/10.5281/zenodo.8077113 can help to get (more) knowledge how to proceed easily and successfully 😉

📖 The manual was developed to help authorities, landscape conservation / maintenance or nature conservation associations etc. in nature conservation to manage their growing monitoring tasks more effectively and cost-efficiently.
🛸 The use of drones can contribute in many ways to increasing the effectiveness of monitoring, reducing costs and minimizing disturbance - https://www.researchgate.net/profile/Steffen-Doering.

Best regards!
Steffen

Esther Githinji

12 August 2024 10:56am

@DibblexLesalon looks like a great question for your team at Expert Drones Africa :) 

See full post
Link

AI Based Animal Detection Demo App for iPhone

Robert (Bob) Zak

11 August 2024 9:49pm

In this post, I describe an animal recognition demonstration app I developed for the iPhone. The “MegaDetector-Demo” app uses the latest “MegaDetector” animal detector model from PyTorch-Wildlife to identify animals, people and vehicles in a live video feed from the iPhone camera. The demo app is available as an Xcode project for Apple developers on my GitHub site. I hope this basic app will encourage more practical uses of edge-based AI tech in trail cameras.

0
discussion

Enhancing Public/Private Safety and Health Through Toilet Mapping

martin wainaina chege

22 July 2024 12:25pm

IntroductionToilets are essential facilities that ensure hygiene and convenience in our communities. However, during periods of heavy rainfall, these vital structures often face...

13 1
See full post
discussion

AI based forest fire early warning system?

Aishwarya Maheshwari

6 August 2024 3:41am

Hello Everyone,Has anyone used or have any experience with forest fire early warning systems? I recently found the "Bosch environmental sensor" Perhaps they claim that it’s the...

2 0
Carly Batist

6 August 2024 3:30pm

Just some additional resources that may be helpful for ya! 

There are some additional start-ups I know of that are working on this, such as Robotics Cats, FireScout, Pano AI, OroraTech, etc. But those are usually image-based (either satellites or drones or static cameras) rather than gas sensors. NOAA is working on a project for this as well (and this). And ConservationX Labs just launched a grand challenge around wildlfire solutions which may be of interest.

See full post
discussion

Spatiotemporal species distribution modeling

Navodita Mathur

6 August 2024 12:39am

Hi,I have a deep passion to contribute in the field of ecology, biodiversity and AI. I have a tech background and about a year of experience in AI. My current focus is on species...

1 0
Will Armstrong

6 August 2024 4:09pm

Hey, 

You're correct, traditionally species distribution models are spatial only, but here are a few ideas for how to incorporate time into your modelling: 

  1. Categorise the time dimension (E.G. Month/Season/Year, wet season/dry season etc) and use it as an input variable. 
  2. Incorporate phenological data (flowering time, migration period etc) into the input variables.
  3. Use time series data and use an appropriate model to take this into account such as LSTM.

A few things to bear in mind:

  1. We need the time of the X data to match that of the y data. This constrains the data that we can use. I'd recommend using Google Earth Engine to get the X data as you can query it for specific date ranges and they have a wide range of relevant datasets. 
  2. Whichever variables you use as input also have to be used for prediction.
  3. Any categorised temporal variables you choose have to be applicable to all species being modelled.

I hope this helps! 

-Will 

See full post
article

Introducing Our Bioacoustics Horizon Scan

Vanesa Reyes

5 August 2024 7:11pm

WILDLABS is thrilled to announce the launch of a Horizon Scan on emerging priorities in Bioacoustics, supported by the Moore Foundation and in collaboration with Colorado State University. This project aims to identify...

2 4
Carly Batist

6 August 2024 3:38pm

Super excited to be in involved in this and can't wait to see what comes of it!!
See full post
discussion

ChatGPT for conservation

Isobel Ashbey

16 January 2023 10:04am

Hi, I've been wondering what this community's thoughts are on ChatGPT? I was just having a play with it and asked:"could you write me a script in python that loads photos and...

53 18
Sara Olsson

17 July 2024 3:23pm

Hi, I've been using ChatGPT for quick labeling of camera trap recordings. I just posted this as a separate discussion here: 

See full post
discussion

efficient methods for labeling clusters of fish in images for segmentation purposes 

Tongfei Yang

16 July 2024 9:43pm

Hi everyone,Our team is working on a project involving the segmentation of individual fish within clusters (groupers specifically). We have images with hundreds of groupers...

1 0
Lampros Mouselimis

5 August 2024 9:58am

hi, if your intention is to use programming and a programming language like Python then one option which can give good results is the SAM (Segment Anything) or SAM-2 algorithm. The following weblinks include more information:

 

See full post
discussion

Rapid Camera Trap labeling with ChatGPT

Sara Olsson

17 July 2024 3:19pm

Hi all, I've published a blog post on rapidly labeling camera trap data using ChatGPT for species identification and a simple object detection model to get the bounding boxes....

3 5
Herdhanu Jayanto

20 July 2024 3:28pm

super interesting, thank you for sharing! definitely will watch!

Sarah Mshanga

3 August 2024 7:46pm

Really interesting, I will take a look

See full post
discussion

Successfully integrated deepfaune into video alerting system

Kim Hendrikse

2 December 2023 11:15am

Hi all, I've successfully integrated deepfaune into my Video alerting full-features security system StalkedByTheState. The yellow box around the image represents the zone of...

20 1
Thijs Suijten

12 May 2024 2:59pm

Yes, this system is designed to be installed near farms. We also have the repeller system with audio & light, that is battery & solar powered. This system is a "last line of defence". The blowers alone requires 1000 watts :)

See full post
discussion

Beginner Bird Classification Project Question

Ashley Kim

1 August 2024 11:56pm

I'm currently working on a bird image classification project for two birds, the warbling vireo and the house wren. I have pulled four months of entries for both of them through...

5 0
Ashley Kim

3 August 2024 1:02am

Hi Carly, thanks for the advice! I hope to use this project only with the two birds using part of the labeled data from iNaturalist, saving the rest to run in the model. I will take a look at Dan's page as well, thank you so much! 

Ashley Kim

3 August 2024 1:05am

Thank you so much Dan! Honestly, "I just want to learn about training image classifiers and I like birds" is a solid characterization of what I'm doing, or at least my initial spark. Although I plan to focus on the two birds for now, I will keep in mind your tips for adding other birds as I try to improve and advance my project. 

Ashley Kim

3 August 2024 1:08am

Hi Carly, thank you for the advice! I am only focusing on the two birds for now - my goal is to use part of the iNaturalist data I have pulled to train the model and the rest to improve it. I will check out Dan's page as well, thanks so much! 

See full post
discussion

AI accelerator for nonprofits working in the Climate area

Angelika Nocon

31 July 2024 3:01pm

Hello everyone! I'm here today to share an interesting opportunity with you! As part of the Tech To The Rescue team, I am thrilled to unveil our latest transformative initiative!...

2 1
Angelika Nocon

1 August 2024 1:39pm

Thank you so much! Now everything is in the hands of amazing organizations and companies! But the first results of the Disaster Management cohort are bringing a very optimistic vision! :) I hope for the same in the Climate cohort!

See full post
discussion

Indigenous communities and AI for Conservation

Sam Reynolds

8 May 2024 12:32pm

Hello! I am looking for recommendations for people from indigenous communities who are either using AI or exploring the potential of AI to solve conservation problems...

6 2
Courtney Shuert

24 May 2024 3:53pm

I am also commenting for future notifications - very interested to hear some responses.

While not directly related to AI, here in Canada there's quite a conversation around data sovereignty for Indigenous communities, such as OCAP (Ownership, Control, Access, Possession) which may be able to connect you with some of the big players in this part of the world. There is also a few efforts to incorporate more Indigenous knowledge systems in statistical modelling which may be of interest: https://www.biorxiv.org/content/10.1101/2023.09.07.556613v1.full.pdf

See full post
discussion

Firetail 12.2 released

Tobias Petri

12 June 2024 7:01pm

Hi all!I'm happy to announce that we just released Firetail - release 12.2A few new things in a nutshellA new evaluation module for evaluating the performance of predicted...

2 2
Tobias Petri

25 June 2024 6:00pm

New short tutorial available



 

See full post
discussion

Mass Detection of Wildlife Snares Using Airborne Synthetic Radar

David Gaynor

7 January 2024 6:50am

Mass Detection of Wildlife Snares Using Airborne Synthetic RadarFor the last year my colleauges Prof. Mike Inggs (Radar - Electrical Engineering, Unviversity of Cape Town) and...

37 7
David Gaynor

14 July 2024 9:54am

Hi Heinrich,  you are absolutely right about the need for long-endurance drone for such a mission on 250km^2.

Taking a full frame 35mm camera with a 35mm (wide angle) lens  you would have a field of view of  123 metres  x 83 metres (looking straight down, you want to be looking at an angle and doing this in high contrast lighting, preferably with the sun behind the camera).

If the area was a square 15.8km x15.8km you would need to do 16 km / 123m = 128 transects to cover the area which is 16km x 128 = 2,057km which if flying at 135 km (safe speed for game counting in a high wing Cessna 182)  is 15.23 hours of flying.

Endurance and speed in commercially available drones - is what is stopping drones from realising their conservation potential.

Attached is initial (conservative)  design performance details.

Contact me on +27 72 472 5318 and we can discuss this further.

 

See full post
discussion

Bringing BearID to Ecuador

Ed Miller

22 July 2024 3:41am

As an employee of Arm, I am able to take a 4 week sabbatical once every 4 years. I used mine to spend a month in Ecuador expanding the BearID Project to a new species, introducing...

1 3
Phil Mitchell

28 July 2024 10:17pm

@bluevalhalla thanks for sharing the awesome video, sounds like a great project!

See full post
discussion

Move BON Development: Follow up discussion

Lacey Hughey

19 June 2024 12:02pm

Hey Biologging Community! We just launched a new initiative to mobilize animal tracking data in support of national and global scale conservation goals (learn more here!). If you...

6 5
Akiba

12 July 2024 5:38am

Hi Talia! 

I feel like the topic is so broad that it might help to put some constraints around things, see what works, and then broaden those out. I have a lot of ideas regarding the data monitoring and collection side based on the other sensor and observation networks we've set up in the past. 

There may also be some potential scope to incorporate things like data collection and integrated monitoring to the Build Your Own Datalogger series where the system is updated to feed data into the observation network. 

It'd probably take a bit of discussion and coordination. Let me know if interested. I'm fine to jump on a call or discuss via email too.

CathyNj

22 July 2024 9:24pm

@cmwainaina please take a look

See full post
Link

Rapid Camera Trap labeling with ChatGPT

Sara Olsson

17 July 2024 3:02pm

Hi all, I've published a blog post on rapidly labeling camera trap data using ChatGPT for species identification and a simple object detection model to get the bounding boxes. While presented in Edge Impulse, this approach can be replicated outside the platform using Python scripts for example. I’d love for you to give it a read and share your thoughts on its potential usefulness in real-world applications.

1