Deep Learning for Fire and Smoke Detection in Outdoor Spaces

Favorskaya, MN; Jain, LC

Deep Learning for Fire and Smoke Detection in Outdoor Spaces

Favorskaya, MN Jain, LC

Permalink

Publisher:: Springer Singapore
Publication Type:: Chapter
Citation:: Smart Innovation, Systems and Technologies, 2021, 215, pp. 195-209
Issue Date:: 2021-01-01

Closed Access

	Filename	Description	Size
	Favorskaya-Jain2021_Chapter_DeepLearningForFireAndSmokeDet.pdf	Published version	416.52 kB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Favorskaya, MN
dc.contributor.author	Jain, LC
dc.date.accessioned	2022-05-15T06:07:54Z
dc.date.available	2022-05-15T06:07:54Z
dc.date.issued	2021-01-01
dc.identifier.citation	Smart Innovation, Systems and Technologies, 2021, 215, pp. 195-209
dc.identifier.isbn	9789813346185
dc.identifier.uri	http://hdl.handle.net/10453/157367
dc.description.abstract	Recent investigations in deep learning provide a new approach for fire and smoke detection in outdoor spaces. Fire and smoke detection is an additional function in smart urban surveillance, as well as, in wildlife monitoring using stationary cameras or UAV cameras. Smoke detection plays an important role in a fire alarm. The algorithms based on the traditional machine learning techniques show high values of errors that cause additional economic costs. Deep learning techniques solve this problem partly but raise new challenges. In this chapter, we analyze deep learning models applicable for this task and present a Weaved Recurrent Single Shot Detector (WRSSD) for early smoke detection with acceptable error ratios.
dc.language	en
dc.publisher	Springer Singapore
dc.relation.ispartof	Smart Innovation, Systems and Technologies
dc.relation.isbasedon	10.1007/978-981-33-4619-2_15
dc.rights	info:eu-repo/semantics/closedAccess
dc.title	Deep Learning for Fire and Smoke Detection in Outdoor Spaces
dc.type	Chapter
utslib.citation.volume	215
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
utslib.copyright.status	closed_access	*
dc.date.updated	2022-05-15T06:07:52Z
pubs.publication-status	Published
pubs.volume	215

Abstract:

Recent investigations in deep learning provide a new approach for fire and smoke detection in outdoor spaces. Fire and smoke detection is an additional function in smart urban surveillance, as well as, in wildlife monitoring using stationary cameras or UAV cameras. Smoke detection plays an important role in a fire alarm. The algorithms based on the traditional machine learning techniques show high values of errors that cause additional economic costs. Deep learning techniques solve this problem partly but raise new challenges. In this chapter, we analyze deep learning models applicable for this task and present a Weaved Recurrent Single Shot Detector (WRSSD) for early smoke detection with acceptable error ratios.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/157367