Modernised Reduction: Adapting the ROT tree

Shane Harrigan; Sonya Coleman; Dermot Kerr; JP Quinn

Modernised Reduction: Adapting the ROT tree

Shane Harrigan, Sonya Coleman, Dermot Kerr, JP Quinn

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

26 Downloads (Pure)

Abstract

Neuromorphic vision data offers a new means of evaluating digitise spatial and temporal representations of evolving scene dynamics. Reduction-Over-Time (ROT) trees have seen growing popularity as a medium for storing and operating over the temporally asynchronous data produced from neuromorphic sensors given their 1-Dnature, spatial preservation, and speed. In this paper we propose a variation of the ROT tree called R-ROT which allows for greater adaptability within structure when compared to the originally proposed ROT tree using adaptive self pruning. The R-ROT structure is evaluated against the original ROT model and is shown to achieve high accuracy results in shorter time across a widely popular benchmark database.

Original language	English
Title of host publication	International Conference on Intelligent Autonomous Systems
Publication status	Accepted/In press - 2023
Event	18th International Conference on Intelligent Autonomous Systems - Suwon, Korea, Democratic People's Republic of Duration: 4 Jul 2023 → 7 Jul 2023 https://ias-18.org/

Conference

Conference	18th International Conference on Intelligent Autonomous Systems
Abbreviated title	IAS18-2023
Country/Territory	Korea, Democratic People's Republic of
City	Suwon
Period	4/07/23 → 7/07/23
Internet address	https://ias-18.org/

Keywords

smart sensors and actuators
smart factory
neuromorphic data

Access to Document

IAS_acceptedAuthor Accepted Manuscript, 439 KB

Cite this

@inproceedings{9ecb19cb4f084faab0fbfbaae7220f70,

title = "Modernised Reduction: Adapting the ROT tree",

abstract = "Neuromorphic vision data offers a new means of evaluating digitise spatial and temporal representations of evolving scene dynamics. Reduction-Over-Time (ROT) trees have seen growing popularity as a medium for storing and operating over the temporally asynchronous data produced from neuromorphic sensors given their 1-Dnature, spatial preservation, and speed. In this paper we propose a variation of the ROT tree called R-ROT which allows for greater adaptability within structure when compared to the originally proposed ROT tree using adaptive self pruning. The R-ROT structure is evaluated against the original ROT model and is shown to achieve high accuracy results in shorter time across a widely popular benchmark database.",

keywords = "smart sensors and actuators, smart factory, neuromorphic data",

author = "Shane Harrigan and Sonya Coleman and Dermot Kerr and JP Quinn",

year = "2023",

language = "English",

booktitle = "International Conference on Intelligent Autonomous Systems",

note = "18th International Conference on Intelligent Autonomous Systems, IAS18-2023 ; Conference date: 04-07-2023 Through 07-07-2023",

url = "https://ias-18.org/",

}

TY - GEN

T1 - Modernised Reduction: Adapting the ROT tree

AU - Harrigan, Shane

AU - Coleman, Sonya

AU - Kerr, Dermot

AU - Quinn, JP

PY - 2023

Y1 - 2023

N2 - Neuromorphic vision data offers a new means of evaluating digitise spatial and temporal representations of evolving scene dynamics. Reduction-Over-Time (ROT) trees have seen growing popularity as a medium for storing and operating over the temporally asynchronous data produced from neuromorphic sensors given their 1-Dnature, spatial preservation, and speed. In this paper we propose a variation of the ROT tree called R-ROT which allows for greater adaptability within structure when compared to the originally proposed ROT tree using adaptive self pruning. The R-ROT structure is evaluated against the original ROT model and is shown to achieve high accuracy results in shorter time across a widely popular benchmark database.

AB - Neuromorphic vision data offers a new means of evaluating digitise spatial and temporal representations of evolving scene dynamics. Reduction-Over-Time (ROT) trees have seen growing popularity as a medium for storing and operating over the temporally asynchronous data produced from neuromorphic sensors given their 1-Dnature, spatial preservation, and speed. In this paper we propose a variation of the ROT tree called R-ROT which allows for greater adaptability within structure when compared to the originally proposed ROT tree using adaptive self pruning. The R-ROT structure is evaluated against the original ROT model and is shown to achieve high accuracy results in shorter time across a widely popular benchmark database.

KW - smart sensors and actuators

KW - smart factory

KW - neuromorphic data

UR - https://ias-18.org/

M3 - Conference contribution

BT - International Conference on Intelligent Autonomous Systems

T2 - 18th International Conference on Intelligent Autonomous Systems

Y2 - 4 July 2023 through 7 July 2023

ER -

Modernised Reduction: Adapting the ROT tree

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this