Saved in:
Bibliographic Details
Main Authors: Lee, Chae Young, Pu, Yi, Fite, Maxwell, Rao, Tejus, Achour, Sara, Kapetanovic, Zerina
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2501.10547
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929681751605248
author Lee, Chae Young
Pu
Yi
Fite, Maxwell
Rao, Tejus
Achour, Sara
Kapetanovic, Zerina
author_facet Lee, Chae Young
Pu
Yi
Fite, Maxwell
Rao, Tejus
Achour, Sara
Kapetanovic, Zerina
contents We present HyperCam, an energy-efficient image classification pipeline that enables computer vision tasks onboard low-power IoT camera systems. HyperCam leverages hyperdimensional computing to perform training and inference efficiently on low-power microcontrollers. We implement a low-power wireless camera platform using off-the-shelf hardware and demonstrate that HyperCam can achieve an accuracy of 93.60%, 84.06%, 92.98%, and 72.79% for MNIST, Fashion-MNIST, Face Detection, and Face Identification tasks, respectively, while significantly outperforming other classifiers in resource efficiency. Specifically, it delivers inference latency of 0.08-0.27s while using 42.91-63.00KB flash memory and 22.25KB RAM at peak. Among other machine learning classifiers such as SVM, xgBoost, MicroNets, MobileNetV3, and MCUNetV3, HyperCam is the only classifier that achieves competitive accuracy while maintaining competitive memory footprint and inference latency that meets the resource requirements of low-power camera systems.
format Preprint
id arxiv_https___arxiv_org_abs_2501_10547
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle HyperCam: Low-Power Onboard Computer Vision for IoT Cameras
Lee, Chae Young
Pu
Yi
Fite, Maxwell
Rao, Tejus
Achour, Sara
Kapetanovic, Zerina
Computer Vision and Pattern Recognition
Machine Learning
Neural and Evolutionary Computing
Image and Video Processing
We present HyperCam, an energy-efficient image classification pipeline that enables computer vision tasks onboard low-power IoT camera systems. HyperCam leverages hyperdimensional computing to perform training and inference efficiently on low-power microcontrollers. We implement a low-power wireless camera platform using off-the-shelf hardware and demonstrate that HyperCam can achieve an accuracy of 93.60%, 84.06%, 92.98%, and 72.79% for MNIST, Fashion-MNIST, Face Detection, and Face Identification tasks, respectively, while significantly outperforming other classifiers in resource efficiency. Specifically, it delivers inference latency of 0.08-0.27s while using 42.91-63.00KB flash memory and 22.25KB RAM at peak. Among other machine learning classifiers such as SVM, xgBoost, MicroNets, MobileNetV3, and MCUNetV3, HyperCam is the only classifier that achieves competitive accuracy while maintaining competitive memory footprint and inference latency that meets the resource requirements of low-power camera systems.
title HyperCam: Low-Power Onboard Computer Vision for IoT Cameras
topic Computer Vision and Pattern Recognition
Machine Learning
Neural and Evolutionary Computing
Image and Video Processing
url https://arxiv.org/abs/2501.10547