Pillar Networks++: Distributed non-parametric deep and wide networks

TL;DR

This paper introduces Pillar Networks++, a distributed non-parametric deep and wide network approach that combines Gaussian Processes with multi-stream CNNs using a Product of Experts framework to improve action recognition without hand-crafted features.

Contribution

It presents a novel integration of Gaussian Processes with deep CNNs in a distributed setting, utilizing a Product of Experts for improved performance.

Findings

Achieved near state-of-the-art accuracy on HMDB-51 dataset.

Eliminated the need for hand-crafted features in action recognition.

Demonstrated effective uncertainty estimation in predictions.

Abstract

In recent work, it was shown that combining multi-kernel based support vector machines (SVMs) can lead to near state-of-the-art performance on an action recognition dataset (HMDB-51 dataset). This was 0.4\% lower than frameworks that used hand-crafted features in addition to the deep convolutional feature extractors. In the present work, we show that combining distributed Gaussian Processes with multi-stream deep convolutional neural networks (CNN) alleviate the need to augment a neural network with hand-crafted features. In contrast to prior work, we treat each deep neural convolutional network as an expert wherein the individual predictions (and their respective uncertainties) are combined into a Product of Experts (PoE) framework.