Deep-plant: Plant identification with convolutional neural networks

This paper studies convolutional neural networks (CNN) to learn unsupervised feature representations for 44 different plant species, collected at the Royal Botanic Gardens, Kew, England. To gain intuition on the chosen features from the CNN model (opposed to a `black box' solution), a visualisa...

Full description

Saved in:
Bibliographic Details
Published in2015 IEEE International Conference on Image Processing (ICIP) pp. 452 - 456
Main Authors Sue Han Lee, Chee Seng Chan, Wilkin, Paul, Remagnino, Paolo
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.09.2015
Subjects
deep learning
Failure analysis
feature visualisation
Machine learning
plant classification
Shape
Support vector machines
Training
Visualization
Yttrium
Online AccessGet full text

Cover

More Information
Summary:This paper studies convolutional neural networks (CNN) to learn unsupervised feature representations for 44 different plant species, collected at the Royal Botanic Gardens, Kew, England. To gain intuition on the chosen features from the CNN model (opposed to a `black box' solution), a visualisation technique based on the deconvolutional networks (DN) is utilized. It is found that venations of different order have been chosen to uniquely represent each of the plant species. Experimental results using these CNN features with different classifiers show consistency and superiority compared to the state-of-the art solutions which rely on hand-crafted features.
DOI:10.1109/ICIP.2015.7350839