To Perceive or Not to Perceive: Lightweight Stacked Hourglass Network

TL;DR

This paper introduces a lightweight stacked hourglass network for human pose estimation that significantly reduces parameters and computational cost while maintaining comparable accuracy.

Contribution

A novel lightweight 2-stacked hourglass network using depthwise separable convolutions and residual connections, with minimal performance loss.

Findings

79% reduction in parameters

Similar performance drop in MAdds

Maintains accuracy with fewer resources

Abstract

Human pose estimation (HPE) is a classical task in computer vision that focuses on representing the orientation of a person by identifying the positions of their joints. We design a lighterversion of the stacked hourglass network with minimal loss in performance of the model. The lightweight 2-stacked hourglass has a reduced number of channels with depthwise separable convolutions, residual connections with concatenation, and residual connections between the necks of the hourglasses. The final model has a marginal drop in performance with 79% reduction in the number of parameters and a similar drop in MAdds