Abstract:In an effort to reduce pesticide use, agronomists and computer scientists have joined forces to develop site-specific weed detection and classification systems. These systems aim to recognize and locate weed species within a crop field, using precision equipment to apply required herbicides timely and only where needed, with the objective of reducing the sprayable surface required to eliminate the given weed and protect the crop, with both economic and environmental benefits. Yet, with climate change on the rise, common weeds are expected to undergo some changes to adapt to their environment, possibly with new or invasive weeds spreading to areas where they did not exist before. These changes (often morphological) as well as new invasions need to be taken into account by future classifiers and detection algorithms to ensure system robustness and adaptation to new habitats/climate dynamics. This paper proposes a set of experiments evaluating the use of transfer learning and zero-shot learning for weed classification using our novel TomatoWeeds dataset. Residual networks of variable depth, pretrained on the Imagenet and/or DeepWeeds datasets were evaluated. A ResNet50 pretrained on both datasets and fine-tuned on the TomatoWeeds dataset performed best, returning a holdout set accuracy of 77.8%, showing the advantageous use of transfer learning in this domain. Zero-shot learning, using both embeddings of images and morphological and habitat text-based descriptions, is implemented to test the ability of machine learning pipelines of recognizing unseen classes at test time (which may arise e.g. due to changing climate dynamics), a learning task in which the field (and our experiments) are still far from satisfactory results. Further research could benefit from larger weed-specific datasets for transfer learning as well as deeper network architectures to improve model performance. The projection-based ZSL could also benefit from larger datasets and new zero-shot learning architectures in hope that unseen classes are accurately projected.

Knowledge Transfer Via Convolution Neural Networks for Multi-Resolution Lawn Weed Classification

Multi-Resolution Weed Classification Via Convolutional Neural Network and Superpixel Based Local Binary Pattern Using Remote Sensing Images.

Transferable Convolutional Neural Network for Weed Mapping with Multisensor Imagery

Analysis of Spectral Bands and Spatial Resolutions for Weed Classification Via Deep Convolutional Neural Network

Deep learning techniques for enhancement of weeds growth classification

Weed Classification in Hyperspectral Remote Sensing Images Via Deep Convolutional Neural Network

Weed Identification by Single-Stage and Two-Stage Neural Networks: A Study on the Impact of Image Resizers and Weights Optimization Algorithms

Weed Density Extraction Based on Few-Shot Learning Through UAV Remote Sensing RGB and Multispectral Images in Ecological Irrigation Area.

Weed recognition using deep learning techniques on class-imbalanced imagery

Evaluation of convolutional neural networks for herbicide susceptibility-based weed detection in turf

CNN feature based graph convolutional network for weed and crop recognition in smart farming

Multi-Class Weed Recognition Using Hybrid CNN-SVM Classifier

Knowledge Distillation and Student-Teacher Learning for Weed Detection in Turf

A deep learning-based method for classification, detection, and localization of weeds in turfgrass.

Transfer and zero-shot learning for scalable weed detection and classification in UAV images

A Deep-Learning- Based Low-Altitude Remote Sensing Algorithm for Weed Classification in Ecological Irrigation Area

Low-Altitude Remote Sensing Based on Convolutional Neural Network for Weed Classification in Ecological Irrigation Area

Deep-CNN based Robotic Multi-Class Under-Canopy Weed Control in Precision Farming

DCNN Transfer Learning and Multi-model Integration for Disease and Weed Identification.

Classification of Weed Seeds Based on Visual Images and Deep Learning