HiRISE: High-Resolution Image Scaling for Edge ML via In-Sensor Compression and Selective ROI

TL;DR

HiRISE is a novel high-resolution image scaling system for edge machine learning that uses in-sensor compression and ROI selection to drastically reduce memory, data transfer, and energy needs on tiny IoT devices.

Contribution

It introduces a new in-sensor image scaling approach with ROI capability, enabling high-resolution image processing on memory-constrained edge devices.

Findings

Up to 17.7x reduction in data transfer and energy consumption.

Significant decrease in peak memory requirements for high-res images.

Effective selective ROI processing on tiny IoT devices.

Abstract

With the rise of tiny IoT devices powered by machine learning (ML), many researchers have directed their focus toward compressing models to fit on tiny edge devices. Recent works have achieved remarkable success in compressing ML models for object detection and image classification on microcontrollers with small memory, e.g., 512kB SRAM. However, there remain many challenges prohibiting the deployment of ML systems that require high-resolution images. Due to fundamental limits in memory capacity for tiny IoT devices, it may be physically impossible to store large images without external hardware. To this end, we propose a high-resolution image scaling system for edge ML, called HiRISE, which is equipped with selective region-of-interest (ROI) capability leveraging analog in-sensor image scaling. Our methodology not only significantly reduces the peak memory requirements, but also achieves up to 17.7x reduction in data transfer and energy consumption.