Deep Predictive Coding Network for Object Recognition

TL;DR

This paper introduces a deep predictive coding network inspired by neuroscience that iteratively refines object recognition through recursive top-down and bottom-up processes, outperforming traditional feedforward models on standard benchmarks.

Contribution

The paper presents a novel bi-directional recurrent neural network architecture based on predictive coding theory, demonstrating improved object recognition performance through recursive inference cycles.

Findings

PCN outperforms feedforward-only models on CIFAR-10/100, SVHN, and MNIST.

Performance improves with more recursive cycles.

PCN can be unfolded into a deeper feedforward model over time.

Abstract

Based on the predictive coding theory in neuroscience, we designed a bi-directional and recurrent neural net, namely deep predictive coding networks (PCN). It has feedforward, feedback, and recurrent connections. Feedback connections from a higher layer carry the prediction of its lower-layer representation; feedforward connections carry the prediction errors to its higher-layer. Given image input, PCN runs recursive cycles of bottom-up and top-down computation to update its internal representations and reduce the difference between bottom-up input and top-down prediction at every layer. After multiple cycles of recursive updating, the representation is used for image classification. With benchmark data (CIFAR-10/100, SVHN, and MNIST), PCN was found to always outperform its feedforward-only counterpart: a model without any mechanism for recurrent dynamics. Its performance tended to improve given more cycles of computation over time. In short, PCN reuses a single architecture to recursively run bottom-up and top-down processes. As a dynamical system, PCN can be unfolded to a feedforward model that becomes deeper and deeper over time, while refining it representation towards more accurate and definitive object recognition.