Deep-Plant: Plant Identification with convolutional neural networks

TL;DR

This paper demonstrates that convolutional neural networks can effectively learn distinctive features for plant identification, outperforming traditional methods by focusing on venation patterns and using visualization techniques for interpretability.

Contribution

The study introduces a CNN-based approach for plant species identification and employs deconvolutional networks for feature visualization, providing insights into the learned features.

Findings

CNN features outperform hand-crafted features in classification accuracy.

Venation patterns are key discriminative features for plant species.

Visualization confirms CNN focuses on venation structures for identification.

Abstract

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.