Hardware-aware mobile building block evaluation for computer vision

TL;DR

This paper introduces a hardware-aware evaluation methodology for neural network building blocks in computer vision, enabling better understanding of accuracy and hardware cost trade-offs across embedded platforms.

Contribution

It presents a novel comparison approach using Pareto fronts for hardware-aware evaluation of neural network building blocks, providing deeper insights into their efficiency.

Findings

Matching previous comparison methods in accuracy insights

Hardware-aware evaluation reveals trade-offs between accuracy and cost

Choosing optimal building blocks can double inference speed on specific hardware

Abstract

In this work we propose a methodology to accurately evaluate and compare the performance of efficient neural network building blocks for computer vision in a hardware-aware manner. Our comparison uses pareto fronts based on randomly sampled networks from a design space to capture the underlying accuracy/complexity trade-offs. We show that our approach allows to match the information obtained by previous comparison paradigms, but provides more insights in the relationship between hardware cost and accuracy. We use our methodology to analyze different building blocks and evaluate their performance on a range of embedded hardware platforms. This highlights the importance of benchmarking building blocks as a preselection step in the design process of a neural network. We show that choosing the right building block can speed up inference by up to a factor of 2x on specific hardware ML accelerators.