WILDLABS logoWildlabs.net : The conservation technology network

WILDLABS logoWildlabs.net : The conservation technology network

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

Help - Innovative ways to track elephant movement

Tyler Nuckols

28 October 2022 4:50pm

Hi, I am working on a study to track wild elephant movement within an overlayed matrix of crop fields. We seek to understand how various landscape variables impact crop-raiding...

7 2
Peter Apps

4 November 2022 5:24pm

Why would you want to avoid alerting the rangers ?

You don't need high tech for this; elephants leave very obvious tracks and sign. 

Kim Lane

7 November 2022 12:52am

Hi Tyler,

Would like to introduce you to Ceres Tags products

  • Ceres Tags products come in boxes of 5, 10 and 24. 
  • There are some software partners such as Earthranger, Mapipedia and possibly CiboLabs that would be able to assist you with your mapping vegetation requirements
  • Ceres Tag does not require any towers, base stations and infrastructure. This allows you to see any movements from the heard outside of their normal herd (boundary alerts), and you will not be disturbing any of the flora and fauna with infrastructure set up. 
  • For the timing you are looking at, Ceres Wild pings directly to satellite 24 times a day. For Ceres Trace and Ceres Ranch there are 4 within 24 hours. Taking into consideration, when you set up alert areas, you will get them directly to your phone/laptop via your software of choice
  • Ceres Ranch is a reusable tag that has just been launched. Use it on this project, remove the tag and then use the tag on your next project
  • The software you choose will assist with the history of your animal movements. Ceres Tag is integrated with 11 software partners and in-development with 18 software partners https://cerestag.com/pages/software-partners
  • Understanding it is a short-term project, you would be able to use Ceres Tags products without the additional expense of setting up and removing infrastructure- towers, gateways
  • With Ceres Tag, you are purchasing the box of tags and picking a suitable software to deliver the information you require. On average, a box of 10 Ceres Trace Tags, is the same as 1 LoRaWAN tower. 
https://www.wildlabs.net/discussion/live-qa-discussion-ceres-tag-animal-monitoring-solution

 

See full post
discussion

CV4Ecology 2023 Applications

Sara Beery

1 November 2022 8:01pm

Hi all! The 2023 CV4Ecology Summer Workshop is open for applications, deadline December 15th 2022.  I'm opening up this thread for potential students to ask questions or find...

5 3
Michael Procko

3 November 2022 7:13pm

Hey Sara,

Thanks again for the session today--it was super helpful! 

I have access to a pretty large network of camera trappers, and I'm currently surveying them to see who might be willing to share data with me. That being said, I can likely get data-sharing agreements with each of these agencies sorted out before the application deadline, but may not have the data in-hand by this time due to a lack of resources for transferring such large amounts of data. 

I've used Azure Storage Explorer before to rapidly transfer large datasets like this, so I think it might be easier to compile all data in one location once we have access to the VMs. Would it be acceptable to apply with the signed data-sharing agreements in mind, and with the intent of organizing all of the data once we acquire access to the computing resources? 

Sara Beery

4 November 2022 6:02pm

Yes, that's definitely ok! You should just outline your larger plan and where you are at in the process in your application. However, the compute resources are intended to be used for model training and won't be available too far in advance of the school (or at least I can't guarantee they will be) , so you would need to make a plan for pre-summer workshop data storage, wrangling, and curation to get the data ready to train models on. Let me know if you're having issues with that and I can try to help brainstorm.

Suzanne Stathatos

9 December 2022 8:23pm

Quick reminder that the deadline for applications is just shy of a week away. 

This workshop is particularly geared to teach ecologists computer vision tools to apply to their research.

Tired of trying to manually count how many animals are in your images? Tired of listening to audio files to classify/detect species? Looking to clean and manage your data so that you can more easily access and analyze it? We're happy and eager to help you apply computer vision tools to try to help with these (and more!) tasks!

More info: 

Feel free to comment here or message me directly if you have any questions!

See full post
discussion

Connection needed!

Alice Marks

15 November 2022 4:23pm

Hi everyone -- I'm hoping to connect with a woman who is based in Asia (South or South East Asia) and is working on AI / machine learning to prevent and reverse environmental...

3 0
Frank van der Most

22 November 2022 2:46pm

Maybe she is not based in Asia, but you could check Lily Xu, who gave a presentation at the IUCN World Congress a year ago on using AI against poaching in Cambodia. The project is called Protection Assistant for Wildlife Security (PAWS).

For a summary of this project and another ( presented by a man ) see 

Read "Machine Learning against poaching and illegal logging – Rubber Boots Data" on Rubber Boots Data

Machine Learning against poaching and illegal logging – Rubber Boots Data

At the IUCN's World Conservation Congress I ran into two applications of machine learning (ML) to guard protected areas against direct detrimental human activity. Both stories are success stories but also both applications are in their early phases. In this post, I will compare the two and ask a couple of common sense questions that…

Rubber Boots Data • 07 Oct 2021
See full post
article

Dear AI, We Have Trust Issues

Dave Thau

14 November 2022 11:09pm

This essay is the third in a series on the promise and challenges of using AI and machine learning to create a planetary environmental management system.  

0
See full post
careers

Spatial Data Analysis, TNC

The Nature Conservancy

8 November 2022 11:34am

Working with a team of conservation practioners, scientists, and technologists, the Spatial Data Analyst uses leading-edge earth observation data in new and effective ways to inform conservation practice locally,...

0
See full post
Link

New paper: Integrating machine learning, remote sensing and citizen science to create an early warning system for biodiversity

Carly Batist

7 November 2022 7:10pm

Opinion paper describing "how data acquired from remote sensing, citizen science & other monitoring approaches could feed in near-real time to an early warning system for biodiversity that integrates automated red-listing of species with the identification of priority areas...

2
https://nph.onlinelibrary.wiley.com/doi/full/10.1002/ppp3.10337
discussion

ConservationAI

ConservationAI

20 October 2022 2:41pm

The conservation AI platform allows you to upload your media for automatic species detection using our trained AI models. The platform uses cutting edge computer vision techniques...

1 1
Carly Batist

20 October 2022 7:32pm

Out of curiosity, what are the similarities/differences between your platform and other image classification ones such as Wildlife Insights, WildID, ZambaCloud? I don't mean that in like a "what's best" way, just would be great to better understand the the details so that people can assess what to try for their specific use cases & applications!

It would be great to have a comparative specs sheet of sorts! Peter Gyurov started something like that: https://www.notion.so/Camera-Trap-Pipeline-Solution-Comparison-2eac80825c4941b0b2b5fad3daea1cc3. Would be great to add ConservationAI to this list!

See full post
article

Audio Across Domains Workshop 2022

Tessa Rhinehart

3 October 2022 7:54pm

A collaborative and cross-disciplinary meeting of audio data scientists spawns creative research collaborations 

2 1
Carly Batist

3 October 2022 11:40pm

Are there any papers you/anyone would recommend discussing or utilizing the representation learning you discuss in the 'Frontiers in data analysis methods' section of your summary...
Justin Kitzes

11 October 2022 2:05pm

Maybe not precisely what you're looking for Carly, but Dan Stowell's new paper in PeerJ is a great introduction to computational methods that are being and could be applied to...
See full post
Link

New paper - Rookognise: Acoustic detection & identification of individual rooks in field recordings using multi-task neural networks

Carly Batist

3 October 2022 2:35pm

Authors developed system to acoustically identify individual animals (under varying conditions) using combo of multi-scale information integration, multi-channel audio & multi-task learning. Tested on captive group of rooks.

0
Link

New paper - FathomNet: A global image database for enabling artificial intelligence in the ocean

Carly Batist

27 September 2022 9:44pm

Open-source image database that standardizes & aggregates expertly curated labeled underwater data, enabling AI in the ocean. Seeded with existing imagery of marine animals, underwater equipment, debris, etc. & allows for future contributions from distributed data...

0
Read "FathomNet: A global image database for enabling artificial intelligence in the ocean | Scientific Reports" on Nature

FathomNet: A global image database for enabling artificial intelligence in the ocean | Scientific Reports

The ocean is experiencing unprecedented rapid change, and visually monitoring marine biota at the spatiotemporal scales needed for responsible stewardship is a formidable task. As baselines are sought by the research community, the volume and rate of this required data collection rapidly outpaces our abilities to process and analyze them. Recent advances in machine learning enables fast, sophisticated analysis of visual data, but have had limited success in the ocean due to lack of data standardization, insufficient formatting, and demand for large, labeled datasets. To address this need, we built FathomNet, an open-source image database that standardizes and aggregates expertly curated labeled data. FathomNet has been seeded with existing iconic and non-iconic imagery of marine animals, underwater equipment, debris, and other concepts, and allows for future contributions from distributed data sources. We demonstrate how FathomNet data can be used to train and deploy models on other institutional video to reduce annotation effort, and enable automated tracking of underwater concepts when integrated with robotic vehicles. As FathomNet continues to grow and incorporate more labeled data from the community, we can accelerate the processing of visual data to achieve a healthy and sustainable global ocean.

Nature
discussion

Conservation Tech Directory - new update!

Carly Batist

29 August 2022 2:38pm

New directory update from @gracieermi & I! And an extra special one as we've just passed our 1-year anniversary! Super exciting to see how far this has come! Check it out:...

2 2
Liz Ferguson

23 September 2022 5:26pm

Congrats on the milestone Carly and Gracie!

See full post