​Recognition inspired by honeybee cognition 

Honeybees recognize flowers with tiny brains, even in noisy conditions. We present a bio-inspired model based on principles observed in the honeybee brain to achieve efficient, multi-task perception under noisy conditions using minimal computational resources.

Trained on synthetic data

Tested on real flowers images

 Demo Task: Flower Color Classification

Watch as the model classifies single flowers by color—just like a bee does!

The model is trained to classify four color classes:

• Utilizes bio-inspired design for energy efficiency
• Transfers knowledge from synthetic data to real-world flowers

• Violet 
• Yellow 
• Orange
• White.

 The Model:

Inspired by the honeybee brain to achieve efficient, multi-task perception in noisy environments.

 ​Low power consumption

 Robust to noise

Bio-inspired design
Trainable parameters in hidden layers 10,504 — a single AI ERN block

Download some sample data