Go to try the model

Video intensity classification with NESA 3x3 

The AI-Lego-Block (ALB) is a bio-inspired AI model based on principles of the honeybee brain, designed for efficient multi-task perception under noisy conditions with minimal computational resources. A live demo is available at ERN-AI Flower Classifier Demo 
ERN-AI© (Evolutionary, Reconfigurable, Neuromorphic-AI) combines multiple micro-data trained ALBs into scalable edge-computing vision systems called Neuromorphic Edge Sensing Arrays (NESA).

NESA Task:

NESA is designed for efficient and accurate motion detection in videos across extreme lighting conditions, from full daylight visibility to ultra-low-light environments with only 2% visible intensity. The system focuses on detecting relevant, unusual, or warning events while minimizing unnecessary processing.

Each ALB specializes in a specific illumination range to improve robustness and detection accuracy:
  • 100% light intensity 
  • 50% light intensity
  • 30% light intensity
  • 15% light intensity
  • 5% light intensity
Together, these specialized ALBs form a modular neuromorphic sensing system capable of adaptive video analysis in real-world edge applications.

Download some sample data

 sample 1
 sample 2
 sample 3

 

Guide tutorial

In progress

Try MODEL HERE