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

Brendan Reidy, Sepehr Tabrizchi, Mohamadreza Mohammadi, Shaahin Angizi, Arman Roohi +1 more
7/23/2024
cs.CV

No official code found. Best implementation: iCAS-Lab/HiRISE-Demo (1 star, MIT, last commit 2025-02-14).

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.

View on arXivView PDF

Code Implementations(1)

iCAS-Lab/HiRISE-DemoCommunity64%
MIT

Demo for the paper entitled "HiRISE: High-Resolution Image Scaling for Edge ML via In-Sensor Compression and Selective ROI" presented at the Design Automation Conference 2024.

10Shell, PythonMay 26, 20241 years agoMIT

Similar Papers

J3DAI: A tiny DNN-Based Edge AI Accelerator for 3D-Stacked CMOS Image Sensor

86% match2025

Ternary-Input Binary-Weight CNN Accelerator Design for Miniature Object Classification System with Query-Driven Spatial DVS

85% match2025

Memory Efficient Continual Learning for Edge-Based Visual Anomaly Detection

84% match2025

Quality Scalable Quantization Methodology for Deep Learning on Edge

84% match2024

MambaLiteSR: Image Super-Resolution with Low-Rank Mamba using Knowledge Distillation

83% match2025

MicroViT: A Vision Transformer with Low Complexity Self Attention for Edge Device

83% match2025

Cite this paper

@article{reidy2024hirise,
  title  = {HiRISE: High-Resolution Image Scaling for Edge ML via In-Sensor Compression and Selective ROI},
  author = {Brendan Reidy and Sepehr Tabrizchi and Mohamadreza Mohammadi and Shaahin Angizi and Arman Roohi and Ramtin Zand},
  year   = {2024},
  eprint = {2408.03956},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2408.03956v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2408.03956.svg)](https://codeofpaper.com/paper/2408.03956)

Discussion