In April of 2024, the BearID Project hosted an AI for Bears Challenge with FruitPunch AI. The challenge was sponsored by Arm and included partners Hack the Planet and NXP. The challenge had two goals:
- Create an efficient classification model for a bear deterrence system
- Create models to detect and identify bears on a “smart camera” device
A total of 19 AI for Good engineers spent more than 1500 hours over 10 weeks working across 4 groups:
- Bear classification
- Bear face detection and segmentation
- Bear identification
- ML pipeline on edge
![](https://bearresearch.org/wp-content/uploads/2024/07/ai4bears-challenge-stats-1024x502.png)
The results were showcased in the Arm booth at Embedded World 2024 in Nuremberg, Germany. For details on the challenge, read the Challenge Result blog on FruitPunch AI.
For a view on why this challenge was important, read the Arm Community blog, AI for Bears Challenge: Empowering developers while fostering conservation.