Tampere University of Technology

TUTCRIS Research Portal

Convolutional low-resolution fine-grained classification

Research output: Contribution to journalArticleScientificpeer-review

Details

Original languageEnglish
Pages (from-to)166-171
JournalPattern Recognition Letters
Volume119
Early online date2017
DOIs
Publication statusPublished - Mar 2019
Publication typeA1 Journal article-refereed

Abstract

Successful fine-grained image classification methods learn subtle details between visually similar (sub-)classes, but the problem becomes significantly more challenging if the details are missing due to low resolution. Encouraged by the recent success of Convolutional Neural Network (CNN) architectures in image classification, we propose a novel resolution-aware deep model which combines convolutional image super-resolution and convolutional fine-grained classification into a single model in an end-to-end manner. Extensive experiments on multiple benchmarks demonstrate that the proposed model consistently performs better than conventional convolutional networks on classifying fine-grained object classes in low-resolution images.

Keywords

  • Deep learning, Fine-grained image classification, Super resolution convoluational neural networks

Publication forum classification

Field of science, Statistics Finland