Schoenhoff, Kurt, Holdsworth, Jason, and Lee, Ickjai (2020) Efficient semantic segmentation through dense upscaling convolutions. In: Proceedings of the 3rd International Conference on Software Engineering and Information Management. pp. 244-248. From: ICSIM'20: 3rd International Conference on Software Engineering and Information Management, 12-15 January 2020, Sydney, NSW, Australia.

PDF (Published Version) - Published Version Restricted to Repository staff only

Abstract

Semantic segmentation is the classification of each pixel in an image to an object, the resultant pixel map has significant usage in many fields. Some fields where this technology is being actively researched is in medicine, agriculture and robotics. For uses where the resources or power requirements are restricted such as robotics or where large amounts of images are required to process, efficiency can be key to the feasibility of a technique. Other applications that require real-time processing have a need for fast and efficient methods, especially where collision avoidance or safety may be involved. We take a combination of existing semantic segmentation methods and improve upon the efficiency by the replacement of the decoder network in ERFNet with a method based upon Dense Upscaling Convolutions, we then add a novel layer that allows the fine tuning of the decoder channel depth and therefore the efficiency of the network. Our proposed modification achieves 20-30% improvement in efficiency on moderate hardware (Nvidia GTX 960) over the original ERFNET and an additional 10% efficiency over the original Dense Upscaling Convolution. We perform a series of experiments to determine viable hyperparameters for the modification and measure the efficiency and accuracy over a range of image sizes, proving the viability of our approach.

Item ID:	63012
Item Type:	Conference Item (Research - E1)
ISBN:	978-1-4503-7690-7
Keywords:	Semantic Segmentation, Classification, Computer Vision, Efficiency, CNN, Deep Learning, Image Processing
Related URLs:	https://dl.acm.org/doi/proceedings/10.11...
Copyright Information:	Copyright © 2020 by the Association for Computing Machinery, Inc. Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from Publications Dept., ACM, Inc., fax +1 (212) 869-0481, or permissions@acm.org.
Date Deposited:	20 Jul 2020 23:39
FoR Codes:	46 INFORMATION AND COMPUTING SCIENCES > 4611 Machine learning > 461103 Deep learning @ 100%
SEO Codes:	89 INFORMATION AND COMMUNICATION SERVICES > 8902 Computer Software and Services > 890299 Computer Software and Services not elsewhere classified @ 100%
Downloads:	Total: 2
	More Statistics