Deep Transfer Learning for UAV-Based Cross-Crop Yield Prediction in Root Crops

Chapter 13 - Update on phytoplasma diseases associated with starchy root and tuber crops in Asia

Sentiment Analysis of Self Driving Car Dataset: A comparative study of Deep Learning approaches

Remote sensing data represent one of the most important sources for automized yield prediction. High temporal and spatial resolution, historical record availability, reliability, and low cost are key factors in predicting yields around the world. Yield prediction as a machine learning task is challenging, as reliable ground truth data are difficult to obtain, especially since new data points can only be acquired once a year during harvest. Factors that influence annual yields are plentiful, and data acquisition can be expensive, as crop-related data often need to be captured by experts or specialized sensors. A solution to both problems can be provided by deep transfer learning based on remote sensing data. Satellite images are free of charge, and transfer learning allows recognition of yield-related patterns within countries where data are plentiful and transfers the knowledge to other domains, thus limiting the number of ground truth observations needed. Within this study, we examine the use of transfer learning for yield prediction, where the data preprocessing towards histograms is unique. We present a deep transfer learning framework for yield prediction and demonstrate its successful application to transfer knowledge gained from US soybean yield prediction to soybean yield prediction within Argentina. We perform a temporal alignment of the two domains and improve transfer learning by applying several transfer learning techniques, such as -SP, BSS, and layer freezing, to overcome catastrophic forgetting and negative transfer problems. Lastly, we exploit spatio-temporal patterns within the data by applying a Gaussian process. We are able to improve the performance of soybean yield prediction in Argentina by a total of 19% in terms of RMSE and 39% in terms of compared to predictions without transfer learning and Gaussian processes. This proof of concept for advanced transfer learning techniques for yield prediction and remote sensing data in the form of histograms can enable successful yield prediction, especially in emerging and developing countries, where reliable data are usually limited.

Emotion recognition (ER) is crucial for understanding human behaviours, social interactions, and psychological well-being. Electroencephalography (EEG) has emerged as a promising tool for capturing the neural correlates of emotions. This work is a systematic review of articles in ER using EEG signals. A total of 120 articles from 1041 articles were selected based on PRISMA guidelines using defined inclusion and exclusion criteria, published between 2018 and 2024. This article aims to provide an in-depth understanding of the current landscape of ER from EEG signals utilizing deep learning (DL). This review offers valuable guidance for researchers and practitioners seeking more refined and reliable emotion classification systems. To explore the effectiveness of DL models in EEG-based ER, several potential DL models, such as convolutional neural network, long short-term memory (LSTM), gated recurrent unit (GRU), hybrid bidirectional LSTM (BiLSTM), bidirectional GRU, and advanced DL models such as convolutional recurrent neural network and EEG-Conformer models are applied to two popular datasets, SEED and GAMEEMO, respectively, to depict the full process of ER. Additionally, the performance of DL models is also compared with the performance of basic machine learning (ML) models such as SVM, k-nearest neighbors, logistic regression, and boosting algorithms such as AdaBoost, XGBoost and LightGBM. Through extensive experiments and performance evaluations, the performance of different models when applied to the datasets mentioned above is compared. The accuracy, precision, recall, and F1-scores are analysed to determine the most effective model for EEG-based ER. The findings of this study demonstrate that the performance of hybrid DL models is more efficacious than that of ML models. The best-performing model (BiLSTM) classified the emotions, with an accuracy of 90.54% when applied to the GAMEEMO dataset. This research contributes to the growing body of literature on ER and provides insights into the feasibility of using EEG signals to understand emotional states, and presents a structured roadmap for future exploration. The findings can aid in the development of more accurate and reliable ER systems, which can have wide-ranging applications in psychology, social sciences, and human-computer interactions.

In today's digital world, app stores have become an essential part of software distribution, providing customers with a wide range of applications and opportunities for software developers to showcase their work. This study elaborates on the importance of end-user feedback for software evolution. However, in the literature, more emphasis has been given to high-rating & popular software apps while ignoring comparatively low-rating apps. Therefore, the proposed approach focuses on end-user reviews collected from 64 low-rated apps representing 14 categories in the Amazon App Store. We critically analyze feedback from low-rating apps and developed a grounded theory to identify various concepts important for software evolution and improving its quality including user interface (UI) and user experience (UX), functionality and features, compatibility and device-specific, performance and stability, customer support and responsiveness and security and privacy issues. Then, using a grounded theory and content analysis approach, a novel research dataset is curated to evaluate the performance of baseline machine learning (ML), and state-of-the-art deep learning (DL) algorithms in automatically classifying end-user feedback into frequently occurring issues. Various natural language processing and feature engineering techniques are utilized for improving and optimizing the performance of ML and DL classifiers. Also, an experimental study comparing various ML and DL algorithms, including multinomial naive Bayes (MNB), logistic regression (LR), random forest (RF), multi-layer perception (MLP), k-nearest neighbors (KNN), AdaBoost, Voting, convolutional neural network (CNN), long short-term memory (LSTM), bidirectional long short term memory (BiLSTM), gated recurrent unit (GRU), bidirectional gated recurrent unit (BiGRU), and recurrent neural network (RNN) classifiers, achieved satisfactory results in classifying end-user feedback to commonly occurring issues. Whereas, MLP, RF, BiGRU, GRU, CNN, LSTM, and Classifiers achieved average accuracies of 94%, 94%, 92%, 91%, 90%, 89%, and 89%, respectively. We employed the SHAP approach to identify the critical features associated with each issue type to enhance the explainability of the classifiers. This research sheds light on areas needing improvement in low-rated apps and opens up new avenues for developers to improve software quality based on user feedback.

The term “root and tuber crops” is a very general “catch-all” for a wide cross-section of subterranean storage organs of which there are approximately 38 root, 23 tuber, 14 rhizome, 11 corm, and 10 bulb crops. The most important among them are potato, sugar beet, carrot, onion, and garlic (temperate root and tuber crops) and cassava, sweet potato, yams, and aroids (tropical root and tuber crops). Pre and postharvest losses of these crops are very high and, depending on the species cultivated and the storage environment, may be of the order of 30–60%. Bacterial control has been emerging as a promising alternative to chemical fungicide to control many pre- and postharvest diseases of these crops. The underlying mechanisms include antibiosis, competition for nutrients and space, and systemic resistance. The various strategies for use of antagonists and field uses (i.e., seed treatment, soil application, foliar spray, antagonistic mixture) have been discussed. In addition, certain mechanisms to improve performance of antagonists, such as addition of nutrients, use of antagonist mixture, and formulation of antagonist have been emphasized in this chapter.

This paper attempts to explore the status of tuber crops cultivation with regard to area, production and productivity across countries and exports of cassava and sweet potatoes from India. The result indicated that among various tuber crops, potatoes were vastly cultivated and consumed by Europe and Asia. At the same time, cassava and sweet potatoes were generally grown and consumed by Africa and Asia. In India, cassava and sweet potatoes are the most important tuber crops due to their large scale and varied uses. The growth rate analysis showed that the area under cassava (-1.38 percent) and sweet potatoes (-0.70 percent) as a whole showed a declining trend in India due to various agro-climatic conditions and socioeconomic constraints. In the context of climate change and considering the importance of root and tuber crops for food and nutritional security, it would be a smart move to bring more area under tuber crops cultivation to achieve 'self-reliance' and ' Make in India Mission'.

Drought effects on root and tuber production: A meta-analysis

Tropical root and tuber crops are gaining more importance presently due to their climate-the word climate resilience is very important for tuber crops resilient nature and irreplaceable role in assuring food security, especially in the developing countries. The major tropical root and tuber crops are cassava/tapioca (Manihot esculenta), sweet potato (Ipomoea batatas), yams (Dioscorea spp.), taro (Colocasia esculenta), elephant foot yam (Amorphophallus paeonifolius) and tania (Xanthosoma sagittifolium). Storage ability of tubers provides an opportunity to market the product, safeguards year-round supply and helps the growers realize better prices. Many viruses and fungi infection of tuber crops cause great loss to growers. The postharvest invasion of various pathogens takes the lead in increasing the loss. Postharvest pathogens make their way to the tuber through wounds or injury that occurs during harvest, transportation, packaging operations and storage processes. The pathogens belonging to the genera Alternaria, Aspergillus, Botrytis, Fusarium, Geotrichum, Gloeosporium, Mucor, Monilinia, Penicillium and Rhizopus cause the most important postharvest diseases in tuber crops. Many physical methods, chemicals, bioagents and botanicals are being utilized for minimizing and managing postharvest diseases. Tuber crops are clonally propagated and thus storage is an essential factor. Minimizing postharvest losses increases food availability to the growing human population, decreases the area needed for production and conserves natural resources. Correct information is essential in formulating an effective and eco-friendly management strategy and thus ensuring tubers that are safe to eat at affordable prices. Precise identification of the pathogens responsible for various diseases, preharvest and postharvest environmental conditions that favour the disease along with proper handling and curing helps in alleviating the huge postharvest loss encountered.

The physicochemical composition and microbial quality of 7 different water sources (Umudike ukwu bore hole (B5), Umuariaga borehole (B3), Nidoro spring water (S2), Umudike bore hole (B4), Umudike school borehole A (B7), Umudike school bore hole B (B6) and Ahiaeke bore hole (B9) used for root and tuber crop processing in Umudike, Nigeria were investigated using standard techniques. The chloride, sulphate, phosphate, dissolved substances, hardness of water, ammonium nitrate and conductivity in all the water samples, fell within the range recommended by the Nigerian Industrial Standard (NIS) or the World Health Organization (WHO) while the values obtained for pH and nitrites were lower than the range given by WHO and NIS. Most of the samples had higher organic carbon contents than permitted. In terms of dissolved oxygen (D) and total hydrogen (TH), S2 had the highest, B5 and B6 had the least for DO (P > 0.05) while B6 had the least in addition for TH. The biochemical oxygen demand (BOD) of B3 was the highest among the water samples analyzed (P 0.05). Sensory evaluation of the samples showed that they were clear, tasteless and odorless. Microbial analysis carried out indicated the presence of coliforms in the water samples although this does not give any cause for concern as they fell within the standards for coliform in water. The total plate count of the samples ranged from 8.00 to 19.09cfu/100ml with B9 having the highest while B7 had the least. Some of the microorganisms isolated include: E.coli, salmonella, shigella, fungi and staphylococcus, indicating microbial contamination of these water samples and thus highlights the need for their urgent sterilization. The absence of Vibrio cholerea in all the water samples indicate the non-prevalence of cholera in the study area. Finally, most of the water sources used by these communities in processing their root and tuber crops did not meet the recommended standards for nitrites, biochemical oxygen demand, organic carbon and nitrites in water, underscoring the need for regular checks and on these water sources. Key words: Water, microbial, coliform, physicochemical, root crops, tuber crops.

Precision genome editing, particularly using Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)/CRISPR-associated protein 9 (Cas9), is advancing crop improvement by enabling targeted and efficient genetic modifications. Root and tuber crops such as potato, cassava, sweet potato, and yam are vital for global food and nutritional security but remain highly vulnerable to climate change, pests, diseases, and limited genetic diversity. Genome editing technologies facilitate the development of improved traits, including enhanced disease resistance, tolerance to abiotic stress, improved nutritional quality, and extended shelf life. This review synthesizes recent advances in genome editing for root and tuber crops across global production systems, including illustrative examples from Sub-Saharan Africa, where active genome editing initiatives are being implemented. It further examines key technical constraints, such as low efficiency of plant transformation and regeneration, and highlights regulatory challenges arising from differing policy frameworks across countries. Emerging solutions are discussed, including genotype-independent editing strategies and DNA-free approaches that avoid the integration of foreign genetic material. Addressing these challenges will be critical for developing resilient and sustainable food systems. Unlike previous reviews, this study integrates mechanistic insights with cross-crop synthesis and proposes next-generation genome editing strategies for engineering complex traits in polyploid root and tuber crops.

The adverse effect of air pollution has always been a problem for human health. The presence of a high level of air pollutants can cause severe illnesses such as emphysema, chronic obstructive pulmonary disease (COPD), or asthma. Air quality prediction helps us to undertake practical action plans for controlling air pollution. The Air Quality Index (AQI) reflects the degree of concentration of pollutants in a locality. The average AQI was calculated for the various cities in China to understand the annual trends. Furthermore, the air quality index has been predicted for ten major cities across China using five different deep learning techniques, namely, Recurrent Neural Network (RNN), Bidirectional Gated Recurrent unit (Bi-GRU), Bidirectional Long Short-Term Memory (BiLSTM), Convolutional Neural Network BiLSTM (CNN-BiLSTM), and Convolutional BiLSTM (Conv1D-BiLSTM). The performance of these models has been compared with a machine learning model, eXtreme Gradient Boosting (XGBoost) to discover the most efficient deep learning model. The results suggest that the machine learning model, XGBoost, outperforms the deep learning models. While Conv1D-BiLSTM and CNN-BiLSTM perform well among the deep learning models in the estimation of the air quality index (AQI), RNN and Bi-GRU are the least performing ones. Thus, both XGBoost and neural network models are capable of capturing the non-linearity present in the dataset with reliable accuracy.

BackgroundPolypyrimidine-tract binding proteins (PTBs) are ubiquitous RNA-binding proteins in plants and animals that play diverse role in RNA metabolic processes. PTB proteins bind to target RNAs through motifs rich in cytosine/uracil residues to fine-tune transcript metabolism. Among tuber and root crops, potato has been widely studied to understand the mobile signals that activate tuber development. Potato PTBs, designated as StPTB1 and StPTB6, function in a long-distance transport system by binding to specific mRNAs (StBEL5 and POTH1) to stabilize them and facilitate their movement from leaf to stolon, the site of tuber induction, where they activate tuber and root growth. Storage tubers and root crops are important sustenance food crops grown throughout the world. Despite the availability of genome sequence for sweet potato, cassava, carrot and sugar beet, the molecular mechanism of root-derived storage organ development remains completely unexplored. Considering the pivotal role of PTBs and their target RNAs in potato storage organ development, we propose that a similar mechanism may be prevalent in storage root crops as well.ResultsThrough a bioinformatics survey utilizing available genome databases, we identify the orthologues of potato PTB proteins and two phloem-mobile RNAs, StBEL5 and POTH1, in five storage root crops - sweet potato, cassava, carrot, radish and sugar beet. Like potato, PTB1/6 type proteins from these storage root crops contain four conserved RNA Recognition Motifs (characteristic of RNA-binding PTBs) in their protein sequences. Further, 3´ UTR (untranslated region) analysis of BEL5 and POTH1 orthologues revealed the presence of several cytosine/uracil motifs, similar to those present in potato StBEL5 and POTH1 RNAs. Using RT-qPCR assays, we verified the presence of these related transcripts in leaf and root tissues of these five storage root crops. Similar to potato, BEL5-, PTB1/6- and POTH1-like orthologue RNAs from the aforementioned storage root crops exhibited differential accumulation patterns in leaf and storage root tissues.ConclusionsOur results suggest that the PTB1/6-like orthologues and their putative targets, BEL5- and POTH1-like mRNAs, from storage root crops could interact physically, similar to that in potato, and potentially, could function as key molecular signals controlling storage organ development in root crops.

Prediction of PM2.5 concentration in urban agglomeration of China by hybrid network model

Objective: To design a deep transfer learning framework for modeling fluence map predictions for stereotactic body radiation therapy (SBRT) of adrenal cancer and similar sites that usually have a small number of cases. Approach: We developed a transfer learning framework for adrenal SBRT planning that leverages knowledge in a pancreas SBRT planning model. Treatment plans from the two sites had different dose prescriptions and beam settings but both prioritized gastrointestinal sparing. A base framework was first trained with 100 pancreas cases. This framework consists of two convolutional neural networks (CNN), which predict individual beam doses (BD-CNN) and fluence maps (FM-CNN) sequentially for 9-beam intensity-modulated radiation therapy (IMRT) plans. Forty-five adrenal plans were split into training/validation/test sets with the ratio of 20/10/15. The base BD-CNN was re-trained with transfer learning using 5/10/15/20 adrenal training cases to produce multiple candidate adrenal BD-CNN models. The base FM-CNN was directly used for adrenal cases. The deep learning (DL) plans were evaluated by several clinically relevant dosimetric endpoints, producing a percentage score relative to the clinical plans. Main results: Transfer learning significantly reduced the number of training cases and training time needed to train such a DL framework. The adrenal transfer learning model trained with 5/10/15/20 cases achieved validation plan scores of 85.4/91.2/90.7/89.4, suggesting that model performance saturated with 10 training cases. Meanwhile, a model using all 20 adrenal training cases without transfer learning only scored 80.5. For the final test set, the 5/10/15/20-case models achieved scores of 73.5/75.3/78.9/83.3. Significance: It is feasible to use deep transfer learning to train an IMRT fluence prediction framework. This technique could adapt to different dose prescriptions and beam configurations. This framework potentially enables DL modeling for clinical sites that have a limited dataset, either due to few cases or due to rapid technology evolution.