Convolutional neural networks (CNNs) have become popular especially in
computer vision in the last few years because they achieved outstanding
performance on different tasks, such as image classifications. We propose a
nine-layer CNN for leaf identification using the famous Flavia and Foliage
datasets. Usually the supervised learning of deep CNNs requires huge datasets
for training. However, the used datasets contain only a few examples per plant
species. Therefore, we apply data augmentation and transfer learning to prevent
our network from overfitting. The trained CNNs achieve recognition rates above
99% on the Flavia and Foliage datasets, and slightly outperform current methods
for leaf classification.
Description
Leaf Identification Using a Deep Convolutional Neural Network
%0 Generic
%1 wick2017leafidentification
%A Wick, Christoph
%A Puppe, Frank
%D 2017
%K cnn dnn myown
%T Leaf Identification Using a Deep Convolutional Neural Network
%U http://arxiv.org/abs/1712.00967
%X Convolutional neural networks (CNNs) have become popular especially in
computer vision in the last few years because they achieved outstanding
performance on different tasks, such as image classifications. We propose a
nine-layer CNN for leaf identification using the famous Flavia and Foliage
datasets. Usually the supervised learning of deep CNNs requires huge datasets
for training. However, the used datasets contain only a few examples per plant
species. Therefore, we apply data augmentation and transfer learning to prevent
our network from overfitting. The trained CNNs achieve recognition rates above
99% on the Flavia and Foliage datasets, and slightly outperform current methods
for leaf classification.
@misc{wick2017leafidentification,
abstract = {Convolutional neural networks (CNNs) have become popular especially in
computer vision in the last few years because they achieved outstanding
performance on different tasks, such as image classifications. We propose a
nine-layer CNN for leaf identification using the famous Flavia and Foliage
datasets. Usually the supervised learning of deep CNNs requires huge datasets
for training. However, the used datasets contain only a few examples per plant
species. Therefore, we apply data augmentation and transfer learning to prevent
our network from overfitting. The trained CNNs achieve recognition rates above
99% on the Flavia and Foliage datasets, and slightly outperform current methods
for leaf classification.},
added-at = {2017-12-18T09:43:34.000+0100},
author = {Wick, Christoph and Puppe, Frank},
biburl = {https://www.bibsonomy.org/bibtex/211e43d7a0432419374e246419b6d52c1/chwick},
description = {Leaf Identification Using a Deep Convolutional Neural Network},
interhash = {8c0a5b250fbd0c4ee7b90dcdbfaf155a},
intrahash = {11e43d7a0432419374e246419b6d52c1},
keywords = {cnn dnn myown},
note = {cite arxiv:1712.00967},
timestamp = {2017-12-18T09:43:34.000+0100},
title = {Leaf Identification Using a Deep Convolutional Neural Network},
url = {http://arxiv.org/abs/1712.00967},
year = 2017
}