# Towards Closing the Energy Gap Between HOG and CNN Features for Embedded Vision

## TL;DR

This paper analyzes the energy, computation, and accuracy trade-offs between CNN and HOG features in embedded vision, aiming to identify how CNNs can become more energy-efficient while maintaining high performance.

## Contribution

It provides a detailed measurement-based comparison of CNN and HOG, offering insights into reducing CNN energy consumption for embedded vision applications.

## Key findings

- Measured energy and accuracy trade-offs on two different chips.
- Identified sources of energy discrepancy between CNN and HOG.
- Suggested potential areas for improving CNN energy efficiency.

## Abstract

Computer vision enables a wide range of applications in robotics/drones, self-driving cars, smart Internet of Things, and portable/wearable electronics. For many of these applications, local embedded processing is preferred due to privacy and/or latency concerns. Accordingly, energy-efficient embedded vision hardware delivering real-time and robust performance is crucial. While deep learning is gaining popularity in several computer vision algorithms, a significant energy consumption difference exists compared to traditional hand-crafted approaches. In this paper, we provide an in-depth analysis of the computation, energy and accuracy trade-offs between learned features such as deep Convolutional Neural Networks (CNN) and hand-crafted features such as Histogram of Oriented Gradients (HOG). This analysis is supported by measurements from two chips that implement these algorithms. Our goal is to understand the source of the energy discrepancy between the two approaches and to provide insight about the potential areas where CNNs can be improved and eventually approach the energy-efficiency of HOG while maintaining its outstanding performance accuracy.