Addax Data Science
Related Inventory Content
AddaxAI is an application designed to streamline the work of ecologists dealing with camera trap images. It’s an AI platform that allows you to analyse images on your local computer and use machine learning models for automatic detection and identification, offering ecologists a way to save time and focus on conservation efforts.
(To avoid any legal concerns, we have renamed our project from EcoAssist to AddaxAI. The project itself remains the same—only the name has changed.)
It has thoe open-source MegaDetector model incorprated, which can filter out images containing animals, people, and vehicles. You can select a species recognition model to further identify the animals. At this moment, there are two species models available: one for Sub-Saharan Africa, and one for Europe. Addax will be adding more identification models in the future. Do you have a model that you would like to make open-source? Or do you want Addax to develop a model specifically for your project? Get in touch!
I'll use this discussion to keep you posted about new models and updates!
Official website: https://addaxdatascience.com/addaxai/
Other WILDLABS threads relating to AddaxAI:
- https://wildlabs.net/article/ecoassist-free-african-species-identification-model-30-species
- https://wildlabs.net/discussion/automatically-detect-animals-camera-trap-images-and-filter-out-empties
11 January 2025 9:35am
Hey @pvlun !
How about support for BioCLIP as a classifier within Eccoassist?
As a Vision Transformer (ViT-B/16) model trained on TreeOfLife-10M, I envision it could be a good starter for annotating new wildlife datasets.
Would it be possible?
Best regards,
Lars
27 January 2025 12:44pm
Can I get access to this model for training datasets in East Africa, Kenya? @pvlun
3 February 2025 1:36pm
Hello, my name is Caroline from Born Free Foundation, Kenya. Do you have more information regarding the Sub-Saharan Africa model? We use Camera Traps for our conservation work in Meru Conservation Area and we struggle a lot with filtering out images containing wildlife, people and vehicles
Lucie Gallegos
Natural Solutions
3 February 2025 9:55pm
Hi @pvlun ,
Thanks for sharing, EcoAssist sounds like a great project, especially with the integration of several models for automatic detection and species identification.
I’m Lucie from Natural Solutions, and we’ve been following your work since 2022. Back then, my colleague @Mathilde worked on creating a global solution for managing data from automated sensors like camera traps (from device management to analysis of the annotation). On our side, we’ve developed an open-source tool called ecoSecrets. We’ve integrated Deepfaune (Europe) for species detection and are working on adding new models from around the world to cover more species.
Our platform also helps to :
- manage device and site info,
- display maps, and
- summarize and analyze annotations (grouped annotations). We’re also working on integrating global biodiversity databases to broaden our species coverage.
We’d love an opportunity to chat to explore potential collaboration and share ideas, especially as we continue expanding our models globally. Looking forward to hearing more about your updates and upcoming models!
Cheers,
Lucie
Lars Holst Hansen
Aarhus University