Using Convolutional Neural Networks to Develop State-of-the-Art Flotation Froth Image Sensors

Convolutional neural networks provide a state-of-the-art approach to the development of froth image sensors. In this study, it is shown that a pretrained neural network architecture, namely VGG16, can be used to obtain significant improvements in froth image sensors. However, training of these netwo...

Full description

Bibliographic Details
Main Authors: Fu, Y., Aldrich, Chris
Format: Journal Article
Published: 2018
Online Access:http://hdl.handle.net/20.500.11937/73068
_version_ 1848762915337797632
author Fu, Y.
Aldrich, Chris
author_facet Fu, Y.
Aldrich, Chris
author_sort Fu, Y.
building Curtin Institutional Repository
collection Online Access
description Convolutional neural networks provide a state-of-the-art approach to the development of froth image sensors. In this study, it is shown that a pretrained neural network architecture, namely VGG16, can be used to obtain significant improvements in froth image sensors. However, training of these networks is computationally demanding and require large data sets that may not be readily available. These problems can be circumvented by making use of transfer learning and partial retraining of the network. Likewise, minor modification of the network architecture can also expedite the development of the models. This is demonstrated in a case study involving an image data set from an industrial platinum group metals plant.
first_indexed 2025-11-14T10:55:09Z
format Journal Article
id curtin-20.500.11937-73068
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T10:55:09Z
publishDate 2018
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-730682019-02-08T07:14:40Z Using Convolutional Neural Networks to Develop State-of-the-Art Flotation Froth Image Sensors Fu, Y. Aldrich, Chris Convolutional neural networks provide a state-of-the-art approach to the development of froth image sensors. In this study, it is shown that a pretrained neural network architecture, namely VGG16, can be used to obtain significant improvements in froth image sensors. However, training of these networks is computationally demanding and require large data sets that may not be readily available. These problems can be circumvented by making use of transfer learning and partial retraining of the network. Likewise, minor modification of the network architecture can also expedite the development of the models. This is demonstrated in a case study involving an image data set from an industrial platinum group metals plant. 2018 Journal Article http://hdl.handle.net/20.500.11937/73068 10.1016/j.ifacol.2018.09.408 restricted
spellingShingle Fu, Y.
Aldrich, Chris
Using Convolutional Neural Networks to Develop State-of-the-Art Flotation Froth Image Sensors
title Using Convolutional Neural Networks to Develop State-of-the-Art Flotation Froth Image Sensors
title_full Using Convolutional Neural Networks to Develop State-of-the-Art Flotation Froth Image Sensors
title_fullStr Using Convolutional Neural Networks to Develop State-of-the-Art Flotation Froth Image Sensors
title_full_unstemmed Using Convolutional Neural Networks to Develop State-of-the-Art Flotation Froth Image Sensors
title_short Using Convolutional Neural Networks to Develop State-of-the-Art Flotation Froth Image Sensors
title_sort using convolutional neural networks to develop state-of-the-art flotation froth image sensors
url http://hdl.handle.net/20.500.11937/73068