Article
Energy & Fuels
Tamara Todic, Vladimir Stankovic, Lina Stankovic
Summary: With the widespread deployment of smart meters, Non-Intrusive Load Monitoring (NILM) has emerged as a promising application for informing energy management within buildings. However, existing deep learning NILM models have limitations in terms of flexibility and scalability. In this study, an active learning framework is proposed to improve transferability and reduce the cost of labeling, achieving optimal accuracy-labelling effort trade-off. The results demonstrate the potential of this approach in improving the performance of NILM models and reducing computational resources needed.
Article
Ecology
Selcan Kaplan Berkaya, Efnan Sora Gunal, Serkan Gunal
Summary: Deep learning-based image classification models are proposed for beehive monitoring, capable of recognizing different conditions and abnormalities with an accuracy of up to 99.07%, making them good candidates for smart beekeeping and beehive monitoring.
ECOLOGICAL INFORMATICS
(2021)
Article
Computer Science, Theory & Methods
Pengzhen Ren, Yun Xiao, Xiaojun Chang, Po-Yao Huang, Zhihui Li, Brij B. Gupta, Xiaojiang Chen, Xin Wang
Summary: Researchers have shown relatively lower interest in active learning compared to deep learning, but with the increasing demand for large-scale high-quality annotated datasets, active learning is receiving more attention. This article provides a comprehensive survey on deep active learning, including a formal classification method, an overview of existing work, and an analysis of developments from an application perspective.
ACM COMPUTING SURVEYS
(2022)
Article
Agriculture, Multidisciplinary
Pieter M. Blok, Gert Kootstra, Hakim Elchaoui Elghor, Boubacar Diallo, Frits K. van Evert, Eldert J. van Henten
Summary: The study aimed to train a CNN with fewer annotated images while maintaining its performance. An active learning method called MaskAL was developed to automatically select hard-to-classify images for annotation and retraining. The results showed that MaskAL outperformed random sampling on a broccoli dataset with visually similar classes.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Agriculture, Multidisciplinary
Cheng-Ju Lee, Ming-Der Yang, Hsin-Hung Tseng, Yu-Chun Hsu, Yu Sung, Wei-Ling Chen
Summary: Single-plant growth monitoring in precision agriculture helps reduce costs and optimize decision-making. This study used UAV imagery and deep learning methods to detect and monitor individual broccoli plants, providing a visualized growth map for precise field management. The proposed approach can be applied to other crops and improve efficiency in agriculture.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2023)
Article
Chemistry, Analytical
Taewon Moon, Dongpil Kim, Sungmin Kwon, Tae In Ahn, Jung Eek Son
Summary: The objective of this study was to establish a non-destructive monitoring system for estimating the fresh weight and leaf area of trellised crops. Using a simple formula and a convolutional neural network, the crop fresh weight and leaf area could be accurately estimated. Therefore, this monitoring system has high versatility and can be widely applied for diverse data analyses.
Article
Engineering, Electrical & Electronic
Nehul Rangappa, Y. Raja Vara Prasad, Shiv Ram Dubey
Summary: Wireless Sensor Networks (WSN) are commonly used for precision agriculture, but reliability and cost are the main limitations. Recently, the use of Unmanned Aerial Vehicles (UAVs) in agriculture has become popular due to their scalability, cost efficiency, and user-friendly adaptations. A novel LED-based wireless communication framework called LEDNet has been proposed, which utilizes LED pattern encoding and computer vision and deep learning techniques for data extraction and communication. The framework shows promising performance in terms of accuracy and power consumption, as observed in experiments with images taken under various lighting and height conditions in the field.
IEEE SENSORS JOURNAL
(2022)
Article
Computer Science, Information Systems
Qilin An, Kai Wang, Zhongyang Li, Chengyuan Song, Xiuying Tang, Jian Song
Summary: This study presents a strawberry growth detection algorithm (SDNet) based on the YOLOX model, which improves the accuracy and speed of strawberry fruit monitoring by enhancing the feature extraction module, attention module, and loss function.
Article
Agriculture, Multidisciplinary
Puneet Singh Thakur, Bhavya Tiwari, Abhishek Kumar, Bhavesh Gedam, Vimal Bhatia, Ondrej Krejcar, Michal Dobrovolny, Jamel Nebhen, Shashi Prakash
Summary: Seed-quality is crucial for achieving uniform seedling growth and high crop yield. In this study, a photonics sensor based on laser backscattering and deep transfer learning is proposed for automatic identification and classification of high-quality seeds. By using convolutional neural networks and transfer learning models, the sensor can extract features from images and achieve high accuracy in seed classification. Experimental results demonstrate the potential of the proposed sensor for accurately monitoring seed quality in real-time applications.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Ecology
Anupong Wongchai, Durga Rao Jenjeti, A. Indira Priyadarsini, Nabamita Deb, Arpit Bhardwaj, Pradeep Tomar
Summary: Agriculture is necessary for human survival, but overpopulation and resource competition pose major challenges to food security. This research proposes a new technique using deep learning for agricultural monitoring and crop disease prediction, which can accurately anticipate the impact of diseases on plants through classification and deep learning.
ECOLOGICAL MODELLING
(2022)
Article
Automation & Control Systems
Jinya Su, Dewei Yi, Baofeng Su, Zhiwen Mi, Cunjia Liu, Xiaoping Hu, Xiangming Xu, Lei Guo, Wen-Hua Chen
Summary: This article explores the use of aerial visual perception for monitoring yellow rust disease, integrating state-of-the-art techniques such as multispectral imaging and deep learning U-Net. The developed framework shows improved performance in disease monitoring by extracting spectral-spatial information simultaneously, surpassing the classical spectral-based classifier (random forest algorithm).
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Agriculture, Multidisciplinary
Stefan Baar, Yosuke Kobayashi, Tatsuro Horie, Kazuhiko Sato, Hidetsugu Suto, Shinya Watanabe
Summary: This paper presents a rail-based video monitoring method for estimating the leaf area index (LAI) of greenhouse tomato plants. The method utilizes optical image segmentation and UNET semantic image segmentation to calculate the relative leaf area over time. The results show that this method performs well with an error of less than ten percent compared to manual estimation, and it is able to accurately distinguish foreground and background plants.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Engineering, Electrical & Electronic
Tanmay Anand, Soumendu Sinha, Murari Mandal, Vinay Chamola, Fei Richard Yu
Summary: Aerial inspection of agricultural regions provides crucial information to safeguard efficient farming, while monitoring farmland anomalies is essential for increasing agricultural technology efficiency and developing AI-assisted farming models. The proposal of the deep learning framework AgriSegNet contributes to automated detection of farmland anomalies and enhancing precision farming techniques.
IEEE SENSORS JOURNAL
(2021)
Article
Agronomy
Andre Silva Aguiar, Sandro Augusto Magalhaes, Filipe Neves dos Santos, Luis Castro, Tatiana Pinho, Joao Valente, Rui Martins, Jose Boaventura-Cunha
Summary: This study used deep learning to detect grape bunches in vineyards, training and deploying two state-of-the-art single-shot multibox models on a low-cost, low-power hardware device to achieve satisfactory performance. Experimental results showed that the models performed better in identifying grape bunches at the medium growth stage.
Article
Plant Sciences
Mingle Xu, Hyongsuk Kim, Jucheng Yang, Alvaro Fuentes, Yao Meng, Sook Yoon, Taehyun Kim, Dong Sun Park
Summary: Recent advancements in deep learning have improved plant disease recognition, but the scarcity of high-quality training datasets hampers their practical application. This paper argues for embracing poor datasets and defines and categorizes the challenges associated with using them. It provides an overview of existing studies and approaches, enhances understanding of these challenges, and contributes to the deployment of deep learning in real-world applications.
FRONTIERS IN PLANT SCIENCE
(2023)