# Unsupervised 3D Object Learning through Neuron Activity aware Plasticity

## TL;DR

This paper introduces a novel neuron activity aware Hebbian learning rule for unsupervised 3D object classification, improving feature retention and accuracy especially with limited training data.

## Contribution

It proposes a dynamic Neuron Activity Aware Hebbian learning rule that enhances unsupervised 3D object learning by balancing neuron activity, outperforming existing Hebbian variants.

## Key findings

- NeAW Hebbian learning reduces neuron activity bias.
- The model achieves higher accuracy than supervised models with limited data.
- Outperforms other Hebbian variants in 3D object classification.

## Abstract

We present an unsupervised deep learning model for 3D object classification. Conventional Hebbian learning, a well-known unsupervised model, suffers from loss of local features leading to reduced performance for tasks with complex geometric objects. We present a deep network with a novel Neuron Activity Aware (NeAW) Hebbian learning rule that dynamically switches the neurons to be governed by Hebbian learning or anti-Hebbian learning, depending on its activity. We analytically show that NeAW Hebbian learning relieves the bias in neuron activity, allowing more neurons to attend to the representation of the 3D objects. Empirical results show that the NeAW Hebbian learning outperforms other variants of Hebbian learning and shows higher accuracy over fully supervised models when training data is limited.