Hasil untuk "Agricultural industries"

Hailiang Zhao, Ziqi Wang, Daojiang Hu et al.

The convergence of artificial intelligence, cyber-physical systems, and cross-enterprise data ecosystems has propelled industrial intelligence to unprecedented scales. Yet, the absence of a unified trust foundation across data, services, and knowledge layers undermines reliability, accountability, and regulatory compliance in real-world deployments. While existing surveys address isolated aspects, such as data governance, service orchestration, and knowledge representation, none provides a holistic, cross-layer perspective on trustworthiness tailored to industrial settings. To bridge this gap, we present \textsc{Trisk} (TRusted Industrial Data-Service-Knowledge governance), a novel conceptual and taxonomic framework for trustworthy industrial intelligence. Grounded in a five-dimensional trust model (quality, security, privacy, fairness, and explainability), \textsc{Trisk} unifies 120+ representative studies along three orthogonal axes: governance scope (data, service, and knowledge), architectural paradigm (centralized, federated, or edge-embedded), and enabling technology (knowledge graphs, zero-trust policies, causal inference, etc.). We systematically analyze how trust propagates across digital layers, identify critical gaps in semantic interoperability, runtime policy enforcement, and operational/information technologies alignment, and evaluate the maturity of current industrial implementations. Finally, we articulate a forward-looking research agenda for Industry 5.0, advocating for an integrated governance fabric that embeds verifiable trust semantics into every layer of the industrial intelligence stack. This survey serves as both a foundational reference for researchers and a practical roadmap for engineers to deploy trustworthy AI in complex and multi-stakeholder environments.

Ziyu Guan, Changhai Qin, Yong Zhao et al.

Accurate estimation of reference crop evapotranspiration (ET0) is crucial for improving water use efficiency and the design and operation of agricultural water management systems. Machine learning (ML) can accurately estimate ET0 across different climatic zones in China when meteorological data are limited, but its “black box” nature restricts interpretability. This study developed an interpretable machine learning workflow to enhance ET0 prediction transparency. It utilized four meta-heuristic algorithms and four machine learning algorithms based on meteorological data observed at 2382 stations across five climatic zones in China from 1960 to 2022. Model performance was evaluated using mean absolute error (MAE), root mean square error (RMSE), Nash-Sutcliffe efficiency coefficient (NSE), coefficient of determination (R2), and global performance index (GPI). Results indicate that the XGBoost model optimized by the Grey Wolf Optimization (GWO) algorithm (GWO-XGB) achieved the highest fitting accuracy. Its Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Nash-Sutcliffe efficiency coefficient (NSE), coefficient of determination (R2), and global prediction index (GPI) were 0.087, 0.116, 0.993, 0.993, and 1.783, respectively. Cross-validation across basins revealed that GWO-XGB maintained an R2 above 0.96 on the independent validation dataset, indicating robust stability and generalization of the interpretable machine learning framework in ET0 prediction. SHAP accurately captured underlying hydrological and climatic processes, identifying solar radiation and extreme temperatures as the primary predictors of ET0, while humidity and wind speed exerted lesser influences. This study offers a promising approach for precise ET0 estimation in data-scarce regions, thereby supporting scientific water resource management and achieving water conservation and efficiency goals. The open-source prediction application is available at https://github.com/guangian.

Shourav Dey, Mohammad Kamrul Hasan, Apurba Adhikary et al.

Early detection of papaya leaf disease is essential for Bangladeshi farmers who experience 39.9% post-harvest losses and pesticide misuse risks. The proposed research presents SwinGhost-ClustNet, which is an ensemble of Swin Transformer to capture the global context, GhostNet to capture the local textures, cross-attention fusion, and K-means clustering (k = 8, silhouette score 0.67) on a 9342-image dataset across eight classes (anthracnose, bacterial spot, healthy, leaf curl, mealybug, mite, mosaic, ring spot) in eight districts.The model reached 99.25% accuracy, 99.28% precision, 99.25% recall, 99.25% F1-score, and 1.000 ROC-AUC on 1401 test images, which is 2.10% higher than base models using pseudo-labels and cross-attention (ablation-validated). Explainable AI (Grad-CAM++ IoU 0.72, Layer-CAM 0.68, and Grad-CAM 0.65) visualizes disease-specific features such as lesions and webbing, and enables building farmer trust.An API based on Flask provides real-time diagnostic, confidence ratings (>70% criterion), and recommendations of pesticides in the Bengali language, which makes it possible to deploy it to the field regardless of single-label restrictions. Future efforts: severity scoring + dosage recommendations, multi-label classification, model pruning (<100 MB), cloud deployment (AWS/GCP), and multi-crop validation with domain adaptation (e.g., 98.89% on lemon leaves; Section 4.7) to address domain shift and species differences. This promotes precision farming, which minimizes losses and excess of chemicals in Bangladesh.

Aliasghar Montazar, John W. Shields, Andre Daccache et al.

Deficit irrigation is increasingly used to reduce agricultural water use in arid regions, yet reliable field-scale quantification of consumptive water savings under commercial conditions remains challenging. Most satellite-based evapotranspiration (ET) studies focus on well-watered systems, whereas deficit irrigation imposes distinct soil–canopy controls on ET response. This study evaluated OpenET satellite-derived actual crop ET (ETc act) for quantifying summer deficit-irrigation impacts in alfalfa systems of California’s Imperial Valley using a two-year field analysis (2024–2025) across 20 grower-managed fields. OpenET ETc act estimates were benchmarked under full irrigation using eddy covariance and under deficit irrigation using SM-RZD. Under full irrigation, OpenET products reproduced daily ETc act dynamics with good agreement, with mean bias errors generally below 0.5 mm d⁻¹ . Following irrigation cutoff, ETc act declined from 6 to 9 mm d⁻¹ to sustained levels near 1–3 mm d⁻¹ , driven primarily by soil-water depletion and reduced transpiration rather than atmospheric demand, supported by declines in NDVI and Kc act. Deficit irrigation reduced cumulative ETc act by 150–200 mm per deficit period, corresponding to 40–50% lower consumptive water use relative to full irrigation (p < 0.001). Despite rainfall variability during the wetter season, deficit-irrigated fields exhibited significantly lower cumulative ET, and the OpenET ensemble provided the most stable representation across conditions. These results indicate that satellite-derived ETc act, when evaluated under water-limited conditions and interpreted using soil-moisture and canopy indicators, provides a credible and operational basis for verifying consumptive water savings from deficit irrigation in arid agricultural systems.

Xiaoyan XIE, Yunxi ZHANG, Ying WANG et al.

Objective: To investigate the in vitro antioxidant activity and stability of white kidney bean antioxidant peptides (WKBAPs). Method: In order to analyze the effects of pH, temperature, metal ions, food ingredient composition, and simulated gastrointestinal digestion on the stability of antioxidant peptides in white kidney beans, the DPPH free radical scavenging rate, ABTS+ free radical scavenging rate, hydroxyl free radical scavenging rate, and total antioxidant capacity were measured, and an oxidative damage model was established in HepG2 cells induced by 2,2'-azodiisobutyramidine dihydrochloride (AAPH) to explore the protective effect of WKBAPs on cellular oxidative damage. Results: Although the antioxidant activity of WKBAPs fluctuated under different processing conditions, the overall performance remained stable, still maintaining a high level of antioxidant activity. Only under strong acid, strong base, high temperatures (>60 ℃), high concentrations of Ca2+, Zn2+, Cu2+, and elevated levels of NaCl, citric acid, and glucose, the antioxidant activity of WKBAPs was significantly diminished (P<0.05). As the K+ concentration increased, the DPPH and ABTS+ radical scavenging capacities of WKBAPs initially decreased and subsequently stabilized. Meanwhile, the hydroxyl radical scavenging capacity and total antioxidant ability exhibited a slight but non-significant decline (P>0.05). During intestinal digestion, the DPPH and hydroxyl radical scavenging rates were significantly reduced, but the high antioxidant activity was still maintained. At the end of 240 minutes intestinal digestion process, its DPPH radical scavenging rate, ABTS+ radical scavenging rate, hydroxyl radical scavenging rate, and total antioxidant capacity were 78.03%±1.01%, 94.05%±0.47%, 74.75%±0.29%, and 0.76±0.02 mmol Fe2+/g, respectively. In the AAPH induced oxidative damage model of HepG2 cells, WKBAPs could increase the survival rate of HepG2 cells damaged by AAP oxidation. It significantly reduced the production of reactive oxygen species (ROS), dramatically decreased the release level of lactate dehydrogenase (LDH) into the cell culture supernatant due to membrane damage, inhibited the production of malondialdehyde (MDA), and enhanced the activity of superoxide dismutase (SOD) and glutathione peroxidase (GSH-Px). Conclusion: WKBAPs has good in vitro antioxidant activity and stability, providing a theoretical basis for its application and development in the field of food and healthcare.

Andreas Mershin, Nikolas Stefanou, Adan Rotteveel et al.

Machine olfaction is rapidly emerging as a transformative capability, with applications spanning non-invasive medical diagnostics, industrial monitoring, agriculture, and security and defense. Recent advances in stabilizing mammalian olfactory receptors and integrating them into biophotonic and bioelectronic systems have enabled detection at near single-molecule resolution thus placing machines on par with trained detection dogs. As this technology converges with multimodal AI and distributed sensor networks imbued with embedded AI, it introduces a new, biochemical layer to a sensing ecosystem currently dominated by machine vision and audition. This review and industry roadmap surveys the scientific foundations, technological frontiers, and strategic applications of machine olfaction making the case that we are currently witnessing the rise of a new industry that brings with it a global chemosensory infrastructure. We cover exemplary industrial, military and consumer applications and address some of the ethical and legal concerns arising. We find that machine olfaction is poised to bring forth a planet-wide molecular awareness tech layer with the potential of spawning vast emerging markets in health, security, and environmental sensing via scent.

Ekin Celikkan, Timo Kunzmann, Yertay Yeskaliyev et al.

Weeds are one of the major reasons for crop yield loss but current weeding practices fail to manage weeds in an efficient and targeted manner. Effective weed management is especially important for crops with high worldwide production such as maize, to maximize crop yield for meeting increasing global demands. Advances in near-sensing and computer vision enable the development of new tools for weed management. Specifically, state-of-the-art segmentation models, coupled with novel sensing technologies, can facilitate timely and accurate weeding and monitoring systems. However, learning-based approaches require annotated data and show a lack of generalization to aerial imaging for different crops. We present a novel dataset for semantic and instance segmentation of crops and weeds in agricultural maize fields. The multispectral UAV-based dataset contains images with RGB, red-edge, and near-infrared bands, a large number of plant instances, dense annotations for maize and four weed classes, and is multitemporal. We provide extensive baseline results for both tasks, including probabilistic methods to quantify prediction uncertainty, improve model calibration, and demonstrate the approach's applicability to out-of-distribution data. The results show the effectiveness of the two additional bands compared to RGB only, and better performance in our target domain than models trained on existing datasets. We hope our dataset advances research on methods and operational systems for fine-grained weed identification, enhancing the robustness and applicability of UAV-based weed management. The dataset and code are available at https://github.com/GFZ/weedsgalore

Vardhan Dongre, Chi Gui, Shubham Garg et al.

We introduce MIRAGE, a new benchmark for multimodal expert-level reasoning and decision-making in consultative interaction settings. Designed for the agriculture domain, MIRAGE captures the full complexity of expert consultations by combining natural user queries, expert-authored responses, and image-based context, offering a high-fidelity benchmark for evaluating models on grounded reasoning, clarification strategies, and long-form generation in a real-world, knowledge-intensive domain. Grounded in over 35,000 real user-expert interactions and curated through a carefully designed multi-step pipeline, MIRAGE spans diverse crop health, pest diagnosis, and crop management scenarios. The benchmark includes more than 7,000 unique biological entities, covering plant species, pests, and diseases, making it one of the most taxonomically diverse benchmarks available for vision-language models, grounded in the real world. Unlike existing benchmarks that rely on well-specified user inputs and closed-set taxonomies, MIRAGE features underspecified, context-rich scenarios with open-world settings, requiring models to infer latent knowledge gaps, handle rare entities, and either proactively guide the interaction or respond. Project Page: https://mirage-benchmark.github.io

Shuangyu Xie, Ken Goldberg, Dezhen Song

Robotic weed removal in precision agriculture introduces a repetitive heterogeneous task planning (RHTP) challenge for a mobile manipulator. RHTP has two unique characteristics: 1) an observe-first-and-manipulate-later (OFML) temporal constraint that forces a unique ordering of two different tasks for each target and 2) energy savings from efficient task collocation to minimize unnecessary movements. RHTP can be framed as a stochastic renewal process. According to the Renewal Reward Theorem, the expected energy usage per task cycle is the long-run average. Traditional task and motion planning focuses on feasibility rather than optimality due to the unknown object and obstacle position prior to execution. However, the known target/obstacle distribution in precision agriculture allows minimizing the expected energy usage. For each instance in this renewal process, we first compute task space partition, a novel data structure that computes all possibilities of task multiplexing and its probabilities with robot reachability. Then we propose a region-based set-coverage problem to formulate the RHTP as a mixed-integer nonlinear programming. We have implemented and solved RHTP using Branch-and-Bound solver. Compared to a baseline in simulations based on real field data, the results suggest a significant improvement in path length, number of robot stops, overall energy usage, and number of replans.

Ali Abdari, Alex Falcon, Giuseppe Serra

Every day, a large amount of educational content is uploaded online across different areas, including agriculture and gardening. When these videos or materials are grouped meaningfully, they can make learning easier and more effective. One promising way to organize and enrich such content is through the Metaverse, which allows users to explore educational experiences in an interactive and immersive environment. However, searching for relevant Metaverse scenarios and finding those matching users' interests remains a challenging task. A first step in this direction has been done recently, but existing datasets are small and not sufficient for training advanced models. In this work, we make two main contributions: first, we introduce a new dataset containing 457 agricultural-themed virtual museums (AgriMuseums), each enriched with textual descriptions; and second, we propose a hierarchical vision-language model to represent and retrieve relevant AgriMuseums using natural language queries. In our experimental setting, the proposed method achieves up to about 62\% R@1 and 78\% MRR, confirming its effectiveness, and it also leads to improvements on existing benchmarks by up to 6\% R@1 and 11\% MRR. Moreover, an extensive evaluation validates our design choices. Code and dataset are available at https://github.com/aliabdari/Agricultural_Metaverse_Retrieval .

Marije Renkema-Singh, Per Hilletofth

Abstract Short food supply chains (SFSCs) have been presented as a sustainable alternative to unsustainable conventional food supply chains (CFSCs). Here, intermediate SFSCs are most capable of expanding their scale and success. However, it remains unclear which intermediaries can do so sustainably, while upholding SFSCs values. Understanding the types of intermediaries and their capabilities is key to ensuring long-term success in the market. Through a systematic literature review, this paper investigates the type of intermediaries operating in SFSCs and their capabilities. Appropriate exclusion criteria eliminated articles that focused on shortening CFSCs without specifically mentioning SFSCs, as well as articles focusing on direct-to-consumer SFSCs, which did not mention intermediaries. A wide range of synonyms for intermediaries and capabilities was included to ensure a thorough search string. Based on these selection criteria, 65 articles have been selected and analyzed. The capabilities of seven different intermediaries are presented in context of the resource-based view. The capabilities are categorized in operational capabilities and relational capabilities. The outcome allows for the assessment of internal resources and external relationships that could strengthen the intermediaries’ operations, ultimately leading to sustainable competitive advantage. The practical implications of this research indicate that certain intermediaries are more suitable than others in obtaining competitive advantage toward CFSC and increasing the volumes moved through SFSCs and allowing for sustainable food supply chains. This is the first review to provide an overview of the most discussed intermediaries in SFSCs and their capabilities creating competitive advantage, allowing for future discussions around sustainable SFSCs.

Osama Magouz, Dina A. Amer, Hanaa A. El-Hamshary et al.

This study investigates the antioxidative and anticancer properties of Salvia officinalis L. hydroethanolic extract, along with its microencapsulation for enhanced delivery and functional fortification in frozen yoghurt. The extract antioxidative activity (0.5 mg/mL) was evaluated using DPPH radical scavenging and ferric reducing power assays, exhibiting 87.67 % scavenging and an increased absorbance of 0.45 compared to the blank sample at 700 nm, respectively. In vitro cytotoxicity (MTT cell viability assay) was assessed against human prostate cancer (PC3), human epidermoid carcinoma (A431), and human hepatocellular carcinoma (HepG2) cell lines, yielding IC50 values of 36.43 µg/mL, 40.78 µg/mL, and 57.45 µg/mL, respectively, with significantly lower cytotoxicity against BJ1 human normal fibroblast cells. Quantitative analysis by high performance liquid chromatography (HPLC-UV) revealed that rosmarinic acid was the extract most abundant phenolic compound, followed by caffeic acid and quercetin. A molecular docking study was subsequently conducted to correlate the in vitro findings, offering insights into key phytochemicals interactions with target proteins. To enhance its bioavailability and stability, Salvia officinalis L. extract was encapsulated via complex coacervation using gum Arabic and whey protein concentrate. The resulting microcapsules were characterized for encapsulation efficiency, morphology, zeta potential, fourier transform infrared (FTIR) spectra, and stability in a simulated gastrointestinal digestion model. These extract-loaded microcapsules were then incorporated into frozen yogurt to improve its functionality. The physicochemical properties, textural profile, viscosity, and sensory evaluation of the fortified frozen yogurt were assessed, demonstrating overall acceptance and enhanced characteristics. These findings highlight the potential of Salvia officinalis L. extract as a bioactive ingredient in dairy products, providing significant antioxidant and anticancer benefits with improved stability and delivery.

Amit Kumar, Jaison Jose, Archit Jain et al.

With recent advancements in industrial robots, educating students in new technologies and preparing them for the future is imperative. However, access to industrial robots for teaching poses challenges, such as the high cost of acquiring these robots, the safety of the operator and the robot, and complicated training material. This paper proposes two low-cost platforms built using open-source tools like Robot Operating System (ROS) and its latest version ROS 2 to help students learn and test algorithms on remotely connected industrial robots. Universal Robotics (UR5) arm and a custom mobile rover were deployed in different life-size testbeds, a greenhouse, and a warehouse to create an Autonomous Agricultural Harvester System (AAHS) and an Autonomous Warehouse Management System (AWMS). These platforms were deployed for a period of 7 months and were tested for their efficacy with 1,433 and 1,312 students, respectively. The hardware used in AAHS and AWMS was controlled remotely for 160 and 355 hours, respectively, by students over a period of 3 months.

Tita Alissa Bach, Aleksandar Babic, Narae Park et al.

The maritime industry requires effective communication among diverse stakeholders to address complex, safety-critical challenges. Industrial AI, including Large Language Models (LLMs), has the potential to augment human experts' workflows in this specialized domain. Our case study investigated the utility of LLMs in drafting replies to stakeholder inquiries and supporting case handlers. We conducted a preliminary study (observations and interviews), a survey, and a text similarity analysis (LLM-as-a-judge and Semantic Embedding Similarity). We discover that while LLM drafts can streamline workflows, they often require significant modifications to meet the specific demands of maritime communications. Though LLMs are not yet mature enough for safety-critical applications without human oversight, they can serve as valuable augmentative tools. Final decision-making thus must remain with human experts. However, by leveraging the strengths of both humans and LLMs, fostering human-AI collaboration, industries can increase efficiency while maintaining high standards of quality and precision tailored to each case.

Thanaporn Singhpoo, Seree Wongpichet, Jetsada Posom et al.

Separating cassava storage root from its stem, known as cassava storage root cutting, represents a pivotal stage in cassava harvesting. It has become increasingly challenging due to a shortage of skilled labor. This research introduces an innovative solution: a cassava storage root-cutting robot (CSRCR) utilizing computer vision technology. The Mask-RCNN model is employed for precise cutting alignment detection. The moving mechanism utilizes a Stewart platform, and the cutting action is performed by a cylinder saw integrated into the robot. The specifications of these components, including dimensions, load capacity, and speed, were meticulously defined and calculated based on a physical survey of cassava plants. The robot's performance was evaluated through a three-step process. First, motion performance was assessed, and the results demonstrated acceptable levels of accuracy, repeatability, and workspace. Second, the optimal moving speed and the cutter's speed were determined. In the third step, the robot was integrated with computer vision technology. The integration achieved a remarkable success rate of 100 %. The average loss and trash were minimized to 1.44 % and 0.66 %, respectively, and the cycle time was 32.43 s. This successful integration not only demonstrates the robot's ability to cut cassava stems accurately in various orientations but also significantly improves efficiency by reducing loss and trash. The research findings pave the way for enhanced traditional cassava harvesting practices.

Василь Суярко, Олег Улицький, Ольга Сердюкова

Statement of the problem determine the factors and processes of contamination of the natural environment with secondary stratum waters. The subject of the research is associated reservoir waters of oil and gas wells. The object of the study is the physical and chemical features of associated reservoir solutions containing hydrocarbons and the processes of their influence on natural ecological systems. The relevance of the research topic is determined by the danger of accompanying reservoir waters and oil products for natural ecological systems and human health. Factors and processes of pollution of the natural environment due to leaks of associated formation water (SW) and oil products in the process of search, exploration and exploitation of hydrocarbon deposits are considered. The mechanisms and ecological consequences of this phenomenon, as well as the level of danger of individual pollutant components, were analyzed. Changes in natural ecosystems as a result of the arrival of sewage sludge and petroleum products into them were studied. Measures to protect the natural environment and reduce environmental risks during drilling of oil and gas wells and development of hydrocarbon deposits are proposed. Accompanying formation waters are brines of sodium chloride solutions (brines) formed at great depths and characterized by high pressures. In the process of drilling deep wells and extracting oil and gas, sewage sludge can gush out onto the earth's surface. This leads to salinization of soils and withdrawal of large areas of agricultural land from circulation, pollution of local water intakes, disruption of the vital activity of flora and fauna, and most importantly – to a negative effect on the human body. It is proposed to divide the types of pollution of sewage containing petroleum products into 4 groups: sodium chloride solutions; petroleum products; toxic components of sewage sludge and petroleum products; thermal pollution by thermal SW. Results. The factors of natural protection of underground water from pollution have been determined. Intensive water exchange in the infiltration zone leads to both rapid pollution and rapid purification of groundwater and non-pressurized groundwater. Underground waters that lie deeper, due to hydrodynamic pressures, push pollutants out of aquifers and remain clean. The chemical elements and compounds of hazardous waste and oil products belonging to the first, second and third classes of danger are given. The consequences of their negative impact on the human body are indicated. A conclusion was made about the need for environmental monitoring near oil and gas industries. The environmental and economic expediency of using SW as a hydro-mineral raw material for the industrial extraction of valuable chemical elements from them is emphasized.

Andrew Katumba, Wayne Steven Okello, Sudi Murindanyi et al.

Beans are essential crops globally, standing out as one of the most consumed and nourishing legumes, thereby playing a significant role in human nutrition and food security. Their cultivation faces several challenges, such as pests, diseases, unpredictable weather patterns, and soil erosion. Of these challenges, diseases are recognized as a key challenge, resulting in a decline in both yield quality and quantity, and inflicting substantial financial losses on farmers.This work proposes a deep learning-based approach for precise in-field identification of diseases in bean plants. We evaluate image classification and object detection models using state-of-the-art Convolutional Neural Network (CNN) architectures to identify Angular Leaf Spot (ALS) and bean rust diseases, key bean diseases in Uganda and the region in general, from smartphone images of bean leaves collected in various districts of Uganda.The dataset employed to train these models is the Makerere University beans image dataset, comprising 15,335 images categorized into three (3) classes (ALS, bean rust, and healthy). To improve in-field performance, the dataset was expanded to include an additional class (unknown class) consisting of a diverse collection of 2,800 images to account for images unrelated to the three (3) predefined classes. Adversarial training was further employed to enhance model robustness in identifying the target classes. In addition, two (2) Out-of-Distribution (ODD) detection techniques, i.e., confidence thresholding and training with an auxiliary class (unknown class), were utilized to handle inputs unrelated to bean leaves.Our results show that our custom CNN, BeanWatchNet, achieved 90% accuracy when tested on unseen data for the classification of the three (3) target classes, i.e., ALS, bean rust and healthy. EfficientNet v2 B0 and BeanWatchNet demonstrated superior performance for the four-class (with unknown class) image classification task, achieving 91% and 90% accuracy, respectively, when evaluated on the test dataset. YOLO v8 exhibited superior performance for the object detection models, attaining mAP@50 of 87.6. The custom CNN model and YOLO v8 model were quantized and deployed across two (2) edge platforms: a smartphone (through a mobile application) and a Raspberry Pi 4B to facilitate in-field disease detection. The benchmarking code and models are publicly available on GitHub.1

Patrick Zimmer, Michael Halstead, Chris McCool

In weed control, precision agriculture can help to greatly reduce the use of herbicides, resulting in both economical and ecological benefits. A key element is the ability to locate and segment all the plants from image data. Modern instance segmentation techniques can achieve this, however, training such systems requires large amounts of hand-labelled data which is expensive and laborious to obtain. Weakly supervised training can help to greatly reduce labelling efforts and costs. We propose panoptic one-click segmentation, an efficient and accurate offline tool to produce pseudo-labels from click inputs which reduces labelling effort. Our approach jointly estimates the pixel-wise location of all N objects in the scene, compared to traditional approaches which iterate independently through all N objects; this greatly reduces training time. Using just 10% of the data to train our panoptic one-click segmentation approach yields 68.1% and 68.8% mean object intersection over union (IoU) on challenging sugar beet and corn image data respectively, providing comparable performance to traditional one-click approaches while being approximately 12 times faster to train. We demonstrate the applicability of our system by generating pseudo-labels from clicks on the remaining 90% of the data. These pseudo-labels are then used to train Mask R-CNN, in a semi-supervised manner, improving the absolute performance (of mean foreground IoU) by 9.4 and 7.9 points for sugar beet and corn data respectively. Finally, we show that our technique can recover missed clicks during annotation outlining a further benefit over traditional approaches.

Silvana Silva

A urbanização ganha novos elementos nas duas primeiras décadas do século XXI. O processo de digitalização, adensamento da camada técnica digital, embora ocorra de maneira desigual no território brasileiro, é um dos elementos transformadores da dinâmica urbana no período recente. A expansão da ação das plataformas digitais, do e-commerce, das redes sociais digitais, sob a hegemonia da racionalidade neoliberal, trouxe mudanças na economia política da cidade e da urbanização. Nesse sentido, buscamos analisar de maneira crítica as particularidades do fenômeno no contexto da formação socioespacial brasileira. Nesta análise, destacamos a importância do neoliberalismo como racionalidade condutora da formação da subjetividade coletiva, cujos princípios da concorrência generalizada, da meritocracia e do individualismo deixam marcas na paisagem e nas sociabilidades urbanas nas cidades brasileiras que examinamos neste artigo. Urbanization has assumed new elements in the first two decades of the 21st century. Even though the digitalization process, the intensification of the digital technical layer, is unevenly taking place in Brazilian territory, it is one of the transforming elements of urban dynamics in the recent period. The widening action of digital platforms, e-commerce, and digital social networks, under the hegemony of neoliberal rationality, has brought about changes in the city’s political economy and urbanization. Accordingly, we critically analyzed the particularities of the phenomenon in the Brazilian socio-spatial formation context. In this analysis, we underline the relevance of neoliberalism as the guiding rationality in collective subjectivity shaping, whose principles of generalized competition, meritocracy, and individualism leave their trace on the landscape and urban sociability in the Brazilian cities we examined in this article.

Ambre Dupuis, Camélia Dadouchi, Bruno Agard

To meet global food requirements while responding to the environmental challenges of the 21st century, an agri-environmental transition towards sustainable agricultural practices is necessary. Crop rotation is an ancestral practice and is a pillar of sustainable agriculture. However, this practice requires more organization on the part of producers for the management of crop inputs. That is why the development of a methodology for forecasting crop rotations in the medium term and at the field level is necessary. However, to date, only a methodology based on the Seq2Seq-LSTM has been theorized without being tested on a concrete case of application. The objective of this article is therefore to evaluate the performance of a Seq2Seq-LSTM methodology to predict crop rotations on a real case. The methodology was applied to a problem of crop rotation prediction for field crop farms in Québec, Canada. Using the Recall(N) metric and a historical sequence of length 6, the next 3 crops grown in a field can be predicted with over 81% success when considering 10 selected options. In addition, the methodology was augmented with contextual information such as economic and meteorological data to refine the forecasts. This augmentation systematically improves the performance of the model. This observation provides a relevant line of research for identifying other factors that influence producers’ decision-making on crop rotation.