dendrocentric-snn-event-classification

name

dendrocentric-snn-event-classification

description

DendroNN dendrocentric neural network methodology for energy-efficient classification of event-based data. Incorporates dendritic computation principles into SNNs for improved spatiotemporal processing. Applies to: event-based vision, neuromorphic computing, dendritic computation, energy-efficient classification. Activation: dendrocentric neural, DendroNN, dendritic computation, event-based classification, dendrite-inspired SNN.

DendroNN: Dendrocentric Neural Networks

Energy-efficient neural networks incorporating dendritic computation principles for classification of event-based (neuromorphic) data.

Metadata

• Source: arXiv:2603.09274
• Published: 2026-03-XX
• Category: cs.LG

Core Methodology

Key Innovation

Integrates dendritic computation — the non-linear processing capabilities of biological neuron dendrites — into artificial neural network architectures for efficient processing of event-based data streams.

Biological Inspiration

• Dendrites perform non-linear spatial and temporal integration of synaptic inputs
• Dendritic branches act as semi-independent computational subunits
• Local dendritic spikes enable hierarchical feature extraction
• Branch-specific plasticity supports efficient learning

Technical Framework

1. Dendritic Compartment Model: Each neuron has multiple dendritic branches with independent processing
2. Event-Based Processing: Designed for neuromorphic sensor data (event cameras, DVS)
3. Energy Efficiency: Sparse event-driven activation with dendritic computation
4. Spatiotemporal Integration: Both spatial (branch-level) and temporal (spike-timing) processing