Our AI

About our AI Models

Conservation AI aims to harness machine learning for various conservation projects. The conservation AI site allows you to upload your media for automatic species detection using our AI models. Using the site you can upload single files or batch upload files to create an online libary of classified data all in one place. The conservation AI platform also supports real-time uploads from devices such as camera traps to enable the real-time detection and classification of animal species.

1 2 ... 7 ►

Our models are trained to classify different species based on their geographical location. To date we have tagged over 200,000 objects across hundreds of different species.

What we can detect

You can use the dashboard below to see if a species is in one of our trained models along with how many images were used during training.

Conservation AI provides a generic detector which can detect Person, Animals and Cars (PAC). If a species is not included in one of our existing models you can select the PAC model to perform basic classification.

Camera Trap Models:

wdt_ID	Scientific Name	Common Name	Model Name	Model Type	Region	Image
2	Equus Quagga	Zebra	Sub Saharan	Vision	Africa
3	Giraffa Camelopardalis	Giraffe	Sub Saharan	Vision	Africa
4	Loxodonta Africana	Elephant	Sub Saharan	Vision	Africa
6	Genus Tragelaphus	Spiral-horned Bovines	Sub Saharan	Vision	Africa
8	Panthera Leo	Lion	Sub Saharan	Vision	Africa

AI Models

• Sub Saharan
• UK Mammals
• UK Birds
• North American Mammals
• Tropical Reefs
• Central Asia Mammals
• Indomalayan Mammals

Model Verson: Sub Saharan V15 (Current)

Train Date:

30/10/2024

Model Changes:

Quality update to reduce missclassifications.

Number of Tags Used in Training:

Model Verson: UK Mammals V10 (Current)

Train Date:

25/07/2024

Model Changes:

Quality update to reduce missclassifications. Changed model architecture.

Number of Tags Used in Training:

 

Model Verson: UK Birds V1

Train Date:

12/05/2021

Model Changes:

Quality update to reduce missclassifications. No new species added.

Number of Tags Used in Training:

Model Verson: North American Mammals V1

Train Date:

25/07/2024

Model Changes:

Quality update to reduce missclassifications. Changed model architecture.

Number of Tags Used in Training:

Tropical Reefs V1(Current)

Train Date:

08/05/2022

Model Changes:

Quality update to reduce missclassifications. No new species added.

Number of Tags Used in Training:

 

Model Verson: Central Asia Mammals V2 (Current)

Train Date:

28/11/2023

Model Changes:

Quality update to reduce missclassifications. New species added: Vulpes Vulpes, Gazella Subgutturosa, Capra Sibirica, Genus Camelus

Number of Tags Used in Training:

Model Verson: Central Asia Mammals V1 (Current)

Train Date:

25/06/2023

Model Changes:

Quality update to reduce missclassifications. New species added:

PantheraUncia, Person.

Number of Tags Used in Training:

Model Verson: Indomalayan Mammals v2 (Current)

Train Date:

30/08/2024

Model Changes:

Quality update to reduce missclassifications. Changed model architecture.

Number of Tags Used in Training: