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

Question About Interpolation

Raul Castillo

18 June 2024 10:52pm

Hello! I am a Computer Science Student experimenting with the multi-step models on TensorFlow's "Time series forecasting" tutorial.I am applying to the models the positional data...

1 0
Larry O'Brien

28 June 2024 6:35pm

The use of synthetic training data such as interpolated sequence values is common, but fraught with the issue of your synthetic signal generating features that are not true to real life. 

Instead, you might think about about appending the values of two long-period sin waves per input element to the sequence going in to your first linear / fully-connected layer. The simplest thing that could possibly work would be to interpret minute of the day and day of the month as the values of your sin wave! (Or perhaps minute/hour if all training sequences are quite short.) Since you’re doing sequence prediction, you would add the appropriate values for each image(?) in the sequence being evaluated.   

With that additional signal going in to the model in the early layers, the NN should have a good chance of learning that the differences in the modulating signal corresponds to distance in time. 

This technigue was popularized by early Large Language Models to encode the distance between words. There’s been refinement (search “Rotary Encoding” for instance) but the basic idea of sin waves generalizes well.

See full post
discussion

Conservation Technology for Human-Wildlife Conflict in Non-Protected Areas: Advice on Generating Evidence

Amit Kaushik

22 January 2024 11:36pm

Hello,I am interested in human-dominated landscapes around protected areas. In my case study, the local community does not get compensation because they are unable to provide...

4 1
Chavoux Luyt

4 February 2024 8:16am

Hi Amit,

The most important thing is that the livestock owners contact you as soon as possible after finding the carcass. We commonly do two things if they contact us on the same day or just after the livestock was killed:

  1. Use CyberTracker (or similar software) on an Android smart phone to record all tracks, bite marks, feeding pattern and any other relevant signs of the reason for the loss with pictures and GPS coordinates. [BTW, Compensation is a big issue -- What do you do if the livestock was stolen? What do you do if a domestic animal killed the livestock? What if it died from disease or natural causes and was scavenged upon by carnivores afterwards?]
  2. In the case of most cats, they would hide the prey (or just mark it by covering it with grass or branches and urinating in the area). In this case you can put up a camera trap on the carcass to capture the animal when it returns to its kill (Reconyx is good if you can afford it - we use mostly Cuddeback with white flash). This will normally only work if the carcass is fresh (so other predators would not be able to smell it and not know where it is yet), so the camera only has to be up for 3-5 days max.

This is not really high-tech, but can be very useful to not only establish which predator was responsible (or if a predator was responsible), but also to record all the evidence for that.

Aaron Kennedy

13 June 2024 8:58pm

Hey Amit, 

This is a great question; from our work, we've seen people do a couple of things. We've even seen people using Ring doorbell footage in urban areas as evidence. 

The best thing we've seen is matching the community needs with existing infrastructure: 

  • Are there existing cameras you can leverage, like the doorbell cameras? 
  • Can public participation monitoring service this, i.e. public submitted photos and videos? 

It also totally depends on the wildlife species you're working with, the interaction, damages, etc. If you've found any good solutions, let me know. I'd love to share that information with our clients here who have constant bear problems. 

 

See full post
discussion

AI Identification Models on Thermal Data

Siddhi Singh

6 June 2024 8:30pm

Hello, my name is Siddhi. I recently joined WildLabs and getting to know different groups. I hope that this is the right place to post! I am interested in AI to identify...

4 0
Siddhi Singh

7 June 2024 4:06am

Wow! This dataset seems great and definitely worth trying out. Do you perhaps have a dataset for deer, elk, and those animals of the sort? 
I live in the mountainous region so deer are very common and easily hit. 

Thank you again

Dan Morris

7 June 2024 5:43pm

Sorry, the only other dataset of thermal camera trap images that I'm aware of is mostly elephants, although it does have some goats:

https://github.com/arribada/human-wildlife-conflict?tab=readme-ov-file#elephant-dataset

See full post
discussion

VIHAR-2024 deadline extension, June 30th (Interspeech satellite event) 

Marius Miron

5 June 2024 10:30am

Dear Wildlabs community,The submission deadline for VIHAR-2024 has been extended to June 30th, 2024. VIHAR-2024 (https://vihar-2024.vihar.org) is the fourth international...

3 0
Alex Rood

5 June 2024 3:05pm

Thanks for sharing this @nkundiushuti ! I think this post would be better suited as an event, that way it will show up on the WILDLABS event calendar page. Let me know if you have any questions on how to make an event post! You just click the +Post button in the top right corner, then click "event."

Marius Miron

7 June 2024 8:29am

hi Alex!! I already posted the event, I just wanted to posted an update: the deadline was extended. 

See full post
discussion

Has anyone combined flying drone surveys with AI for counting wild herds?

Johnathan

14 April 2024 3:40pm

My vision is a drone, the kind that fly a survey pattern. Modern ones have a 61 megapixel camera and LIDAR which is a few millimeter resolution, they are for mapping before a road...

24 0
Kim Hendrikse

27 May 2024 10:55am

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.

Jared Marley

31 May 2024 5:47pm

Hi Johnathan, 



There is a Canadian company more or less doing that. They have their own endurance drone and optical/thermal cameras. Very much keyed into surveys and they may have success given the number of helicopter accidents we have had in Western Canada. Not sure if the AI part is there yet. 



I know they've done surveys with at least one department here but not much beyond that. I talked to one of the developers their just as a point of interest. The current leadership today looks different than I remember though. 



 

Johnathan

6 June 2024 2:48pm

The camera can be aimed at the greenhouse background, which is like a huge green screen. Inside the greenhouse there's only a few flying insects, and they would all have to fly between the optics and the wall or roof eventually. Or if the bot is flying, have it look upwards. 

It's pretty much a programing question. Unfortunately I am not the type of person who is good at both building and troubleshooting hardware, and writing code. I took some programming back in college but I am not sure if I want to get myself up to speed. It's starting to sound like I need a few years of college before I can even get started. Which I already did, too bad none of it counts for anything anymore. Or I guess I can compete in the marketplace with people with real money behind them, which is the only thing that means anything. If you are brilliant and not funded, you might as well be a scarecrow.

See full post
article

New WILDLABS Funding & Finance group

Frank van der Most

5 June 2024 3:24pm

WildLabs will soon launch a 'Funding and Finance' group. What would be your wish list for such a group? Would you be interested in co-managing or otherwise helping out?

3 2
Alex Rood

5 June 2024 4:14pm

This is great, Frank! @StephODonnell, maybe we can try to bring someone from #Superorganism (@tomquigley ?) or another venture company (#XPRIZE) into the fold!
Paul Edonga

6 June 2024 1:38am

I find the group to be dope, fundraising in the realm of conservation has been tough especially for emerging conservation leaders. There are no centralized grants tracking common...
See full post
event

Bioacoustics and AI 101

Vanesa Reyes

29 May 2024 2:21pm

Recent developments in AI have not only led to dramatic increase in accuracy of detecting/classifying sounds, but have simultaneously made these tools accessible for people with little to no prior knowledge of AI. This...

0
See full post
event

5th World Ecoacoustics Congress

Vanesa Reyes

28 May 2024 8:07pm

In 2024, the 5th WEC, which will be held from 8th to 12th July in Madrid, will aim to bring together cutting-edge researchers and the international scientific community around this emerging field. During five days, we...

0
See full post
Link

DeepDive: estimating global biodiversity patterns through time using deep learning

Alex Rood

20 May 2024 4:51pm

These authors "develop an approach based on stochastic simulations of biodiversity and a deep learning model to infer richness at global or regional scales through time while incorporating spatial, temporal and taxonomic sampling variation."

0