Skip to main content
DORAS
DCU Online Research Access Service
Login (DCU Staff Only)
FastSal: a computationally efficient network for visual saliency prediction

Hu, Feiyan ORCID: 0000-0001-7451-6438 and McGuinness, Kevin ORCID: 0000-0003-1336-6477 (2021) FastSal: a computationally efficient network for visual saliency prediction. In: 25th International Conference on Pattern Recognition (ICPR2020), 10-15 Jan 2021, Milan, Italy (Online). (In Press)

Full text available as:

[img]
Preview
PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
939kB

Abstract

This paper focuses on the problem of visual saliency prediction, predicting regions of an image that tend to attract hu- man visual attention, under a constrained computational budget. We modify and test various recent efficient convolutional neural network architectures like EfficientNet and MobileNetV2 and compare them with existing state-of-the-art saliency models such as SalGAN and DeepGaze II both in terms of standard accuracy metrics like Area Under Curve (AUC) and Normalized Scanpath Saliency (NSS), and in terms of the computational complexity and model size. We find that MobileNetV2 makes an excellent backbone for a visual saliency model and can be effective even without a complex decoder. We also show that knowledge transfer from a more computationally expensive model like DeepGaze II can be achieved via pseudo-labelling an unlabelled dataset, and that this approach gives result on-par with many state-of-the-art algorithms with a fraction of the computational cost and model size.

Item Type:Conference or Workshop Item (Paper)
Event Type:Conference
Refereed:Yes
Uncontrolled Keywords:Saliency
Subjects:Computer Science > Artificial intelligence
Computer Science > Image processing
Computer Science > Machine learning
Engineering > Signal processing
DCU Faculties and Centres:DCU Faculties and Schools > Faculty of Engineering and Computing > School of Electronic Engineering
Research Initiatives and Centres > INSIGHT Centre for Data Analytics
Publisher:Institute of Electrical and Electronics Engineers (IEEE)
Official URL:http://dx.doi.org/
Copyright Information:© 2020 The Authors
Funders:Science Foundation Ireland (SFI) under grant number SFI/15/SIRG/3283 and SFI/12/RC/2289 P2.
ID Code:25180
Deposited On:17 Nov 2020 13:00 by Feiyan Hu . Last Modified 11 Jan 2021 11:59

Downloads

Downloads per month over past year

Archive Staff Only: edit this record

Altmetric
- Altmetric
+ Altmetric
  • Student Email
  • Staff Email
  • Student Apps
  • Staff Apps
  • Loop
  • Disclaimer
  • Privacy
  • Contact Us