Transfer learning approach for classification and noise reduction on noisy web data

•Gathering an unsupervised image dataset of vehicle models from a marketing site.•Studying the behavior of six deep networks using the transfer learning method.•Discarding the external noise of the dataset using Isolation Forest method.•Demonstrating that a high recognition accuracy is attainable wi...

Full description

Saved in:
Bibliographic Details
Published in:Expert systems with applications 2018-09, Vol.105, p.221-232
Main Authors: Abbasi Aghamaleki, Javad, Moayed Baharlou, Sina
Format: Article
Language:English
Subjects:
Artificial intelligence
Artificial neural networks
Classification schemes
Convolutional Neural Networks
Datasets
Isolation forest
Neural networks
Noise control
Noise reduction
Noisy dataset
Recognition
Training
Transfer learning
Vehicle recognition
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•Gathering an unsupervised image dataset of vehicle models from a marketing site.•Studying the behavior of six deep networks using the transfer learning method.•Discarding the external noise of the dataset using Isolation Forest method.•Demonstrating that a high recognition accuracy is attainable with the given method.•Providing an android application in which the proposed approach is implemented. One of the main ingredients to learn a visual representation of an object using the Convolutional Neural Networks is a large and carefully annotated dataset. Acquiring a dataset in a demanded scale is not a straightforward task; therefore, the community attempts to solve this problem by creating noisy datasets gathered from web sources. In this paper, this issue is tackled by designing a vehicle recognition system using Convolutional Neural Networks and noisy web data. In the proposed system, the transfer learning technique is employed, and behavior of several deep architectures trained on a noisy dataset are studied. In addition, the external noise of the gathered dataset is reduced by exploiting an unsupervised method called Isolation Forest, and the new training results are examined. Based on the experiments, high recognition accuracies were achieved by training two states of the art networks on the noisy dataset, and the obtained results were slightly improved by using the proposed noise reduction framework. Finally, a demonstration application is provided to show the capability and the performance of the proposed approach.
ISSN:0957-4174
1873-6793
DOI:10.1016/j.eswa.2018.03.042