MOSAIC: Spatially-multiplexed edge AI optimization over multiple concurrent video sensing streams
Sustaining high fidelity and high throughput of perception tasks over vision sensor streams on edge devices remains a formidable challenge, especially given the continuing increase in image sizes (e.g., generated by 4K cameras) and complexity of DNN models. One promising approach involves criticalit...
Saved in:
Main Authors: | , , , , |
---|---|
Format: | text |
Language: | English |
Published: |
Institutional Knowledge at Singapore Management University
2023
|
Subjects: | |
Online Access: | https://ink.library.smu.edu.sg/sis_research/7886 https://ink.library.smu.edu.sg/context/sis_research/article/8892/viewcontent/MOSAIC_MMSys23_Camera_ready.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Singapore Management University |
Language: | English |
id |
sg-smu-ink.sis_research-8892 |
---|---|
record_format |
dspace |
spelling |
sg-smu-ink.sis_research-88922024-09-03T07:27:51Z MOSAIC: Spatially-multiplexed edge AI optimization over multiple concurrent video sensing streams GOKARN, Ila SABBELLA, Hemanth HU, Yigong ABDELZAHER, Tarek MISRA, Archan Sustaining high fidelity and high throughput of perception tasks over vision sensor streams on edge devices remains a formidable challenge, especially given the continuing increase in image sizes (e.g., generated by 4K cameras) and complexity of DNN models. One promising approach involves criticality-aware processing, where the computation is directed selectively to "critical" portions of individual image frames. We introduce MOSAIC, a novel system for such criticality-aware concurrent processing of multiple vision sensing streams that provides a multiplicative increase in the achievable throughput with negligible loss in perception fidelity. MOSAIC determines critical regions from images received from multiple vision sensors and spatially bin-packs these regions using a novel multi-scale Mosaic Across Scales (MoS) tiling strategy into a single `canvas frame', sized such that the edge device can retain sufficiently high processing throughput. Experimental studies using benchmark datasets for two tasks, Automatic License Plate Recognition and Drone-based Pedestrian Detection, shows that MOSAIC, executing on a Jetson TX2 edge device, can provide dramatic gains in the throughput vs. fidelity tradeoff. For instance, for drone-based pedestrian detection, for a batch size of 4, MOSAIC can pack input frames from 6 cameras to achieve (a) 4.75X (475%) higher throughput (23 FPS per camera, cumulatively 138FPS) with ≤ 1% accuracy loss, compared to a First Come First Serve (FCFS) processing paradigm. 2023-06-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/7886 info:doi/10.1145/3587819.3590986 https://ink.library.smu.edu.sg/context/sis_research/article/8892/viewcontent/MOSAIC_MMSys23_Camera_ready.pdf http://creativecommons.org/licenses/by/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Edge AI Machine Perception Canvas-based Processing Artificial Intelligence and Robotics Data Science Graphics and Human Computer Interfaces |
institution |
Singapore Management University |
building |
SMU Libraries |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
SMU Libraries |
collection |
InK@SMU |
language |
English |
topic |
Edge AI Machine Perception Canvas-based Processing Artificial Intelligence and Robotics Data Science Graphics and Human Computer Interfaces |
spellingShingle |
Edge AI Machine Perception Canvas-based Processing Artificial Intelligence and Robotics Data Science Graphics and Human Computer Interfaces GOKARN, Ila SABBELLA, Hemanth HU, Yigong ABDELZAHER, Tarek MISRA, Archan MOSAIC: Spatially-multiplexed edge AI optimization over multiple concurrent video sensing streams |
description |
Sustaining high fidelity and high throughput of perception tasks over vision sensor streams on edge devices remains a formidable challenge, especially given the continuing increase in image sizes (e.g., generated by 4K cameras) and complexity of DNN models. One promising approach involves criticality-aware processing, where the computation is directed selectively to "critical" portions of individual image frames. We introduce MOSAIC, a novel system for such criticality-aware concurrent processing of multiple vision sensing streams that provides a multiplicative increase in the achievable throughput with negligible loss in perception fidelity. MOSAIC determines critical regions from images received from multiple vision sensors and spatially bin-packs these regions using a novel multi-scale Mosaic Across Scales (MoS) tiling strategy into a single `canvas frame', sized such that the edge device can retain sufficiently high processing throughput. Experimental studies using benchmark datasets for two tasks, Automatic License Plate Recognition and Drone-based Pedestrian Detection, shows that MOSAIC, executing on a Jetson TX2 edge device, can provide dramatic gains in the throughput vs. fidelity tradeoff. For instance, for drone-based pedestrian detection, for a batch size of 4, MOSAIC can pack input frames from 6 cameras to achieve (a) 4.75X (475%) higher throughput (23 FPS per camera, cumulatively 138FPS) with ≤ 1% accuracy loss, compared to a First Come First Serve (FCFS) processing paradigm. |
format |
text |
author |
GOKARN, Ila SABBELLA, Hemanth HU, Yigong ABDELZAHER, Tarek MISRA, Archan |
author_facet |
GOKARN, Ila SABBELLA, Hemanth HU, Yigong ABDELZAHER, Tarek MISRA, Archan |
author_sort |
GOKARN, Ila |
title |
MOSAIC: Spatially-multiplexed edge AI optimization over multiple concurrent video sensing streams |
title_short |
MOSAIC: Spatially-multiplexed edge AI optimization over multiple concurrent video sensing streams |
title_full |
MOSAIC: Spatially-multiplexed edge AI optimization over multiple concurrent video sensing streams |
title_fullStr |
MOSAIC: Spatially-multiplexed edge AI optimization over multiple concurrent video sensing streams |
title_full_unstemmed |
MOSAIC: Spatially-multiplexed edge AI optimization over multiple concurrent video sensing streams |
title_sort |
mosaic: spatially-multiplexed edge ai optimization over multiple concurrent video sensing streams |
publisher |
Institutional Knowledge at Singapore Management University |
publishDate |
2023 |
url |
https://ink.library.smu.edu.sg/sis_research/7886 https://ink.library.smu.edu.sg/context/sis_research/article/8892/viewcontent/MOSAIC_MMSys23_Camera_ready.pdf |
_version_ |
1814047851832934400 |