Saved Queries

by Mohamed Farag Taha, Hanping Mao, Zhao Zhang, Gamal Elmasry, Mohamed A. Awad, Alwaseela Abdalla, Samar Mousa, Abdallah Elshawadfy Elwakeel and Osama Elsherbiny

Agriculture 2025, 15(6), 582; https://doi.org/10.3390/agriculture15060582 (registering DOI) - 9 Mar 2025

Abstract

Agriculture 5.0 (Ag5.0) represents a groundbreaking shift in agricultural practices, addressing the global food security challenge by integrating cutting-edge technologies such as artificial intelligence (AI), machine learning (ML), robotics, and big data analytics. To adopt the transition to Ag5.0, this paper comprehensively reviews the role of AI, machine learning (ML) and other emerging technologies to overcome current and future crop management challenges. Crop management has progressed significantly from early agricultural methods to the advanced capabilities of Ag5.0, marking a notable leap in precision agriculture. Emerging technologies such as collaborative robots, 6G, digital twins, the Internet of Things (IoT), blockchain, cloud computing, and quantum technologies are central to this evolution. The paper also highlights how machine learning and modern agricultural tools are improving the way we perceive, analyze, and manage crop growth. Additionally, it explores real-world case studies showcasing the application of machine learning and deep learning in crop monitoring. Innovations in smart sensors, AI-based robotics, and advanced communication systems are driving the next phase of agricultural digitalization and decision-making. The paper addresses the opportunities and challenges that come with adopting Ag5.0, emphasizing the transformative potential of these technologies in improving agricultural productivity and tackling global food security issues. Finally, as Agriculture 5.0 is the future of agriculture, we highlight future trends and research needs such as multidisciplinary approaches, regional adaptation, and advancements in AI and robotics. Ag5.0 represents a paradigm shift towards precision crop management, fostering sustainable, data-driven farming systems that optimize productivity while minimizing environmental impact. Full article

(This article belongs to the Special Issue Computational, AI and IT Solutions Helping Agriculture)

24 pages, 3801 KiB

Open AccessArticle

Risk-Adjusted Performance of Random Forest Models in High-Frequency Trading

by Akash Deep, Abootaleb Shirvani, Chris Monico, Svetlozar Rachev and Frank Fabozzi

J. Risk Financial Manag. 2025, 18(3), 142; https://doi.org/10.3390/jrfm18030142 (registering DOI) - 9 Mar 2025

Abstract

Because of the theoretical challenges posed by the Efficient Market Hypothesis with respect to technical analysis, the effectiveness of technical indicators in high-frequency trading remains inadequately explored, particularly at the minute-level frequency, where the effects of the microstructure of the market dominate. This study evaluates the integration of traditional technical indicators with Random Forest regression models using minute-level SPY data, analyzing 13 distinct model configurations. Our empirical results reveal a stark contrast between in-sample and out-of-sample performance, with

R^{2}

values deteriorating from 0.749–0.812 during training to negative values in testing. A feature importance analysis demonstrates that primary price-based features dominate the predictions made by the model, accounting for over 60% of the importance, while established technical indicators, such as RSI and Bollinger Bands, account for only 14–15%. Although the indicator-enhanced models achieved superior risk-adjusted metrics, with Rachev ratios between 0.919 and 0.961, they consistently underperformed a simple buy-and-hold strategy, generating returns ranging from −2.4% to −3.9%. These findings challenge conventional assumptions about the usefulness of technical indicators in algorithmic trading, suggesting that in high-frequency contexts, they may be more relevant to risk management rather than to predicting returns. For practitioners and researchers, our findings indicate that successful high-frequency trading strategies should focus on adaptive feature selection and regime-specific modeling rather than relying on traditional technical indicators, as well as indicating the critical importance of robust out-of-sample testing in the development of a model. Full article

(This article belongs to the Special Issue Machine Learning Applications in Finance, 2nd Edition)

►▼ Show Figures

Figure 1

28 pages, 13449 KiB

Open AccessArticle

DUIncoder: Learning to Detect Driving Under the Influence Behaviors from Various Normal Driving Data

by Haoran Zhou, Alexander Carballo, Masaki Yamaoka, Minori Yamataka, Keisuke Fujii and Kazuya Takeda

Sensors 2025, 25(6), 1699; https://doi.org/10.3390/s25061699 (registering DOI) - 9 Mar 2025

Abstract

Driving Under the Influence (DUI) has emerged as a significant threat to public safety in recent years. Despite substantial efforts to effectively detect DUI, the inherent risks associated with acquiring DUI-related data pose challenges in meeting the data requirements for training. To address this issue, we propose DUIncoder, which is an unsupervised framework designed to learn exclusively from normal driving data across diverse scenarios to detect DUI behaviors and provide explanatory insights. DUIncoder aims to address the challenge of collecting DUI data by leveraging diverse normal driving data, which can be readily and continuously obtained from daily driving. Experiments on simulator data show that DUIncoder achieves detection performance superior to that of supervised learning methods which require additional DUI data. Moreover, its generalization capabilities and adaptability to incremental data demonstrate its potential for enhanced real-world applicability. Full article

(This article belongs to the Special Issue Advanced Sensing and Analysis Technology in Transportation Safety)

31 pages, 5731 KiB

Open AccessArticle

Curriculum-Guided Adversarial Learning for Enhanced Robustness in 3D Object Detection

by Jinzhe Huang, Yiyuan Xie, Zhuang Chen and Ye Su

Sensors 2025, 25(6), 1697; https://doi.org/10.3390/s25061697 (registering DOI) - 9 Mar 2025

Abstract

The pursuit of robust 3D object detection has emerged as a critical focus within the realm of computer vision. This paper presents a curriculum-guided adversarial learning (CGAL) framework, which significantly enhances the adversarial robustness and detection accuracy of the LiDAR-based 3D object detector PointPillars. By employing adversarial learning with prior curriculum expertise, this framework effectively resists adversarial perturbations generated by a novel attack method, P-FGSM, on 3D point clouds. By masterfully constructing a nonlinear enhancement block (NEB) based on the radial basis function network for PointPillars to adapt to the CGAL, a novel 3D object detector named Pillar-RBFN was developed; it exhibits intrinsic adversarial robustness without undergoing adversarial training. In order to tackle the class imbalance issue within the KITTI dataset, a data augmentation technique has been designed that singly samples the point cloud with additional ground truth objects frame by frame (SFGTS), resulting in the creation of an adversarial version of the original KITTI dataset named Adv-KITTI. Moreover, to further alleviate this issue, an adaptive variant of focal loss was formulated, effectively directing the model’s attention to challenging objects during the training process. Extensive experiments demonstrate that the proposed CGAL achieves an improvement of

0.8 \sim 2.5

percentage points in mean average precision (mAP) compared to conventional training methods, and the models trained with Adv-KITTI have shown an enhancement of at least 15 percentage points in mAP, compellingly testifying to the effectiveness of our method. Full article

(This article belongs to the Section Sensing and Imaging)

31 pages, 866 KiB

Open AccessReview

Pharmaceutical 3D Printing Technology Integrating Nanomaterials and Nanodevices for Precision Neurological Therapies

by Jurga Bernatoniene, Mindaugas Plieskis and Kestutis Petrikonis

Pharmaceutics 2025, 17(3), 352; https://doi.org/10.3390/pharmaceutics17030352 (registering DOI) - 9 Mar 2025

Abstract

Pharmaceutical 3D printing, combined with nanomaterials and nanodevices, presents a transformative approach to precision medicine for treating neurological diseases. This technology enables the creation of tailored dosage forms with controlled release profiles, enhancing drug delivery across the blood−brain barrier (BBB). The integration of nanoparticles, such as poly lactic-co-glycolic acid (PLGA), chitosan, and metallic nanomaterials, into 3D-printed scaffolds improves treatment efficacy by providing targeted and prolonged drug release. Recent advances have demonstrated the potential of these systems in treating conditions like Parkinson’s disease, epilepsy, and brain tumors. Moreover, 3D printing allows for multi-drug combinations and personalized formulations that adapt to individual patient needs. Novel drug delivery approaches, including stimuli-responsive systems, on-demand dosing, and theragnostics, provide new possibilities for the real-time monitoring and treatment of neurological disorders. Despite these innovations, challenges remain in terms of scalability, regulatory approval, and long-term safety. The future perspectives of this technology suggest its potential to revolutionize neurological treatments by offering patient-specific therapies, improved drug penetration, and enhanced treatment outcomes. This review discusses the current state, applications, and transformative potential of 3D printing and nanotechnology in neurological treatment, highlighting the need for further research to overcome the existing challenges. Full article

(This article belongs to the Special Issue Applications of Nanomaterials in Drug Delivery and Drug Release)

27 pages, 9185 KiB

Open AccessArticle

Fault Diagnosis of Hydro-Turbine Based on CEEMDAN-MPE Preprocessing Combined with CPO-BILSTM Modelling

by Nengpeng Duan, Yun Zeng, Fang Dao, Shuxian Xu and Xianglong Luo

Energies 2025, 18(6), 1342; https://doi.org/10.3390/en18061342 (registering DOI) - 9 Mar 2025

Abstract

The accuracy of hydro-turbine fault diagnosis directly impacts the safety and operational efficiency of hydroelectric power generation systems. This paper addresses the challenge of low diagnostic accuracy in traditional methods under complex environments. This is achieved by proposing a signal preprocessing method that combines complete ensemble empirical mode decomposition with adaptive noise and multiscale permutation entropy (CEEMDAN-MPE) and that is optimized with the crested porcupine optimizer algorithm for the bidirectional long- and short-term memory network (CPO-BILSTM) model for hydro-turbine fault diagnosis. The method performs signal denoising using CEEMDAN, while MPE extracts key features. Furthermore, the hyperparameters of the CPO-optimized BILSTM model are innovatively introduced. The extracted signal features are fed into the CPO-BILSTM model for fault diagnosis. A total of 150 sets of acoustic vibrational signals are collected for validation using the hydro-turbine test bench under different operating conditions. The experimental results demonstrate that the diagnostic accuracy of the method is 96.67%, representing improvements of 23.34%, 16.67%, and 6.67% over traditional models such as LSTM (73.33%), CNN (80%), and BILSTM (90%), respectively. In order to verify the effectiveness of the signal preprocessing method, in this paper, the original signal, the signal processed by CEEMDAN, CEEMDAN-PE, and CEEMDAN-MPE are input into the CPO-BILSTM model for controlled experiments. The results demonstrate that CEEMDAN-MPE effectively denoises hydro-turbine acoustic vibrational signals while preserving key features. The method in this paper integrates signal preprocessing and deep learning models and, with the help of intelligent optimization algorithms, significantly enhances the model’s adaptive ability, improves the model’s applicability under complex operating conditions, and provides a valuable supplement for hydro-turbine fault diagnosis. Full article

(This article belongs to the Section A3: Wind, Wave and Tidal Energy)

►▼ Show Figures

Figure 1

15 pages, 1808 KiB

Open AccessArticle

The Role of the Abacus and Physical Exercise in the Cognitive Development of Students in Primary Education

by María del Carmen Carcelén-Fraile, Agustín Aibar-Almazán, José Luis Solas-Martínez and Vânia Loureiro

Educ. Sci. 2025, 15(3), 335; https://doi.org/10.3390/educsci15030335 (registering DOI) - 9 Mar 2025

Abstract

(1) Background: Cognitive stimulation during the first years of school is key to the comprehensive development of children, as it impacts functions such as attention, memory, and intelligence, and contributes to their academic performance and social adaptation. The present study aims to evaluate how the use of the abacus and physical exercise improve cognitive skills in children in the second year of primary school. (2) Methods: This study is a randomized clinical trial with a total of 82 children, of which 58.50% were boys and 41.50% girls in the first cycle of primary education, divided into an experimental group that carried out a combined program of training with abacus and physical exercise and a control group. Selective attention and concentration were measured with the D2 test, memory with the Spanish adaptation of the Reynolds Intelligence Scale, differential perception with the Differential Perception Test (CARA-R), and general intelligence with the Raven Progressive Matrices Test. (3) Results: The intervention showed statistically significant improvements in attention (Cohen’s d = 0.55), concentration (Cohen’s d = 0.04), memory (Cohen’s d = 0.53), differential perception (Cohen’s d = 0.77), impulsivity control (Cohen’s d = 0.90), and general intelligence (Cohen’s d = 0.43) within the experimental group, as well as significant differences between the training and control groups in post-intervention assessments. (4) Conclusions: The combination of physical exercise and abacus training effectively improves children’s cognitive development. Full article

(This article belongs to the Special Issue Physical Activity and Active Proposals as a Pathway to Enhance Intelligence and Cognitive/Academic Variables in Young People)

►▼ Show Figures

Figure 1

33 pages, 13441 KiB

Open AccessArticle

DMDiff: A Dual-Branch Multimodal Conditional Guided Diffusion Model for Cloud Removal Through SAR-Optical Data Fusion

by Wenjuan Zhang, Junlin Mei and Yuxi Wang

Remote Sens. 2025, 17(6), 965; https://doi.org/10.3390/rs17060965 (registering DOI) - 9 Mar 2025

Abstract

Optical remote sensing images, as a significant data source for Earth observation, are often impacted by cloud cover, which severely limits their widespread application in Earth sciences. Synthetic aperture radar (SAR), with its all-weather, all-day observation capabilities, serves as a valuable auxiliary data source for cloud removal (CR) tasks. Despite substantial progress in deep learning (DL)-based CR methods utilizing SAR data in recent years, challenges remain in preserving fine texture details and maintaining image visual authenticity. To address these limitations, this study proposes a novel diffusion-based CR method called the Dual-branch Multimodal Conditional Guided Diffusion Model (DMDiff). Considering the intrinsic differences in data characteristics between SAR and optical images, we design a dual-branch feature extraction architecture to enable adaptive feature extraction based on the characteristics of the data. Then, a cross-attention mechanism is employed to achieve deep fusion of the multimodal feature extracted above, effectively guiding the progressive diffusion process to restore cloud-covered regions in optical images. Furthermore, we propose an image adaptive prediction (IAP) strategy within the diffusion model, specifically tailored to the characteristics of remote sensing data, which achieves a nearly 20 dB improvement in PSNR compared to the traditional noise prediction (NP) strategy. Extensive experiments on the airborne, WHU-OPT-SAR, and LuojiaSET-OSFCR datasets demonstrate that DMDiff outperforms SOTA methods in terms of both signal fidelity and visual perceptual quality. Specifically, on the LuojiaSET-OSFCR dataset, our method achieves a remarkable 17% reduction in the FID metric over the second-best method, while also yielding significant enhancements in quality assessment metrics such as PSNR and SSIM. Full article

►▼ Show Figures

Figure 1

24 pages, 2406 KiB

Open AccessReview

A Systematic Review and Meta-Analysis of Effect of Post-Activation Performance Enhancement in Combat Sports: A Systematic Review and Meta-Analysis—Part I: General Performance Indicators

by Artur Terbalyan, Karol Skotniczny, Michał Krzysztofik, Jakub Chycki, Vadim Kasparov and Robert Roczniok

J. Funct. Morphol. Kinesiol. 2025, 10(1), 88; https://doi.org/10.3390/jfmk10010088 (registering DOI) - 9 Mar 2025

Abstract

Background/Objectives: Post-activation performance enhancement (PAPE) has been explored for its potential to improve general performance in combat sports. This systematic review and meta-analysis investigated the effects of PAPE protocols on physical performance, focusing on differences across disciplines, competitive levels, and testing methods. Methods: A PRISMA-guided search (2010–2023) identified 19 studies examining PAPE protocols in combat sports athletes. The inclusion criteria required human trials using defined PAPE protocols, with outcomes of general performance indicators such as countermovement jumps (CMJs). A meta-analysis was conducted on data from 866 athletes using random effects modeling. Results: The PAPE protocols yielded a pooled effect size of 0.136 (95% CI, 0.008–0.263) across 866 athletes. Taekwondo athletes exhibited the most pronounced improvements in CMJ performance, particularly when using protocols that combined repeated vertical jumps with heavy-resistance cluster sets, and with dynamic, sport-specific movements such as the bandal chagui protocol achieving an effect size of 1.19 (p < 0.001). Conversely, Muay Thai athletes experienced performance declines when the protocols incorporated highly specific techniques, such as roundhouse kicks (ES = −1.36, p = 0.009). Analysis by competitive level revealed pooled effect sizes of 0.14 (95% CI, −0.01 to 0.29) for amateur athletes and 0.13 (95% CI, −0.11 to 0.38) for elite athletes, with no statistically significant differences observed between these groups. Conclusions: PAPE’s effectiveness depends on tailoring protocols to the competitive level and discipline. Short rest intervals support plyometric protocols for amateurs, while heavy-resistance exercises enhance elite performers. Further research is needed to standardize PAPE protocols and explore discipline-specific adaptations. Full article

(This article belongs to the Special Issue Optimizing Post-activation Performance Enhancement)

►▼ Show Figures

Figure 1

20 pages, 8765 KiB

Open AccessArticle

Mamba-DQN: Adaptively Tunes Visual SLAM Parameters Based on Historical Observation DQN

by Xubo Ma, Chuhua Huang, Xin Huang and Wangping Wu

Appl. Sci. 2025, 15(6), 2950; https://doi.org/10.3390/app15062950 (registering DOI) - 9 Mar 2025

Abstract

The parameter configuration of traditional visual SLAM algorithms usually relies on expert experience and extensive experiments, and the parameter configuration needs to be reset as the scene changes, which is a complex and tedious process. To achieve parameter adaptation in visual SLAM, we propose the Mamba-DQN method, which transforms complex parameter adjustment tasks into policy learning assignments for the agent. In this paper, we select the key parameters of visual SLAM to construct the agent action space. The reward function is constructed based on the absolute trajectory error (ATE), and the Mamba history observer is built within the agent to learn the observation trajectory, aiming to improve the quality of the agent’s decisions. Finally, the proposed method was experimented on the EuRoc MAV and TUM-VI datasets. The experimental results show that Mamba-DQN not only enhances the positioning accuracy of visual SLAM and demonstrates good real-time performance but also avoids the tedious parameter adjustment process. Full article

(This article belongs to the Special Issue Applications in Computer Vision and Image Processing)

14 pages, 2697 KiB

Open AccessArticle

Seasonal Activity Patterns of Captive Arabian Sand Gazelle (Gazella marica, Thomas, 1897) in Qatar

by Nima Mahmoud, Romaan Hayat Khattak and Muhammad Ali Nawaz

Animals 2025, 15(6), 778; https://doi.org/10.3390/ani15060778 (registering DOI) - 9 Mar 2025

Abstract

The Arabian sand gazelle (Gazella marica) is a native and highly adaptable species of the Arabian Peninsula. Due to drastic population declines, the species is listed as globally vulnerable. Very little is known about the behavioral ecology of this species in captivity; therefore, this study was designed to investigate the seasonal variations in the activity patterns of Arabian sand gazelles at Al Reem Biosphere Reserve, Qatar. Data were collected in two phases, i.e., summer (September–October 2021) and winter (December 2021–January 2022), for a total of 16 days. Results revealed that feeding and walking (p = 0.001) were the dominant activities in both seasons, yet these were higher in summer compared to winter. Likewise, standing, lying down and other activities (social interactions, defecating, maintenance, sexual behaviors) were also higher in summer compared to winter. All these findings suggest that Arabian sand gazelles are adaptable to harsh environments. However, we strongly recommend a year-round investigation on the impacts of humans, feed types and Arabian Oryx on the behavioral activities of Arabian sand gazelles. In addition, we suggest studying the behavior ecology of the wild scattered populations of Arabian sand gazelles for better management of captive breeding stocks. Full article

(This article belongs to the Section Wildlife)

►▼ Show Figures

Figure 1

25 pages, 6769 KiB

Open AccessArticle

NursingXR: Advancing Nursing Education Through Virtual Reality-Based Training

by Mohammad F. Obeid, Ahmed Ewais and Mohammad R. Asia

Appl. Sci. 2025, 15(6), 2949; https://doi.org/10.3390/app15062949 (registering DOI) - 9 Mar 2025

Abstract

The increasing complexity of healthcare delivery and the advancements in medical technology have highlighted the necessity for improved training in nursing education. While traditional training methods have their merits, they often encounter challenges such as limited access to clinical placements, static physical simulations, and performance anxiety during hands-on practice. Virtual reality (VR) has been increasingly adopted for immersive and interactive training environments, allowing nursing students to practice essential skills repeatedly in realistic, risk-free settings. This study presents NursingXR, a VR-based platform designed to help nursing students master essential clinical skills. With a scalable and flexible architecture, NursingXR is tailored to support a variety of nursing lessons and adapt to evolving curricula. The platform has a modular design and offers two interactive modes: Training Mode, which provides step-by-step guided instruction, and Evaluation Mode, which allows for independent performance assessment. This article details the development process of the platform, including key design principles, system architecture, and implementation strategies, while emphasizing its utility and scalability. A mixed-methods evaluation involving 78 participants—both novices and experts—was conducted to evaluate the platform’s usability and user satisfaction. The results underscore NursingXR’s effectiveness in fostering an effective and engaging learning environment as well as its potential as a supplementary resource for nursing training. Full article

(This article belongs to the Special Issue Virtual and Augmented Reality: Theory, Methods, and Applications)

►▼ Show Figures

Figure 1

47 pages, 16513 KiB

Open AccessArticle

Adaptive Sliding Mode Control of an Interleaved Buck Converter–Proton Exchange Membrane Electrolyzer for a Green Hydrogen Production System

by Mohamed Koundi, Hassan El Fadil, Abdellah Lassioui and Yassine El Asri

Processes 2025, 13(3), 795; https://doi.org/10.3390/pr13030795 (registering DOI) - 9 Mar 2025

Abstract

This paper presents an advanced Adaptive Sliding Mode Control (ASMC) strategy, specifically developed for a hydrogen production system based on a Proton Exchange Membrane electrolyzer (PEM electrolyzer). This work utilized a static model of the PEM electrolyzer, characterized by its V-I electrical characteristic, which was approximated by a linear equation. The ASMC was designed to estimate the coefficients of this equation, which are essential for designing an efficient controller. The primary objective of the proposed control strategy is to ensure the overall stability of the integrated system comprising both an interleaved buck converter (IBC) and PEM electrolyzer. The control framework aims to maintain the electrolyzer voltage at its reference value despite the unknown coefficients while ensuring equal current distribution among the three parallel legs of the IBC. The effectiveness of the proposed approach was demonstrated through numerical simulations in MATLAB-SIMULINK and was validated by the experimental results. The results showed that the proposed ASMC achieved a voltage tracking error of less than 2% and a current distribution imbalance of only 1.5%. Furthermore, the controller exhibited strong robustness to parameter variations, effectively handling fluctuations in the electrolyzer’s ohmic resistance (R_ohm) (from ±28.75% to ±40.35%) and in the reversible voltage (E_rev) (from ±28.67% to ±40.19%), highlighting its precision and reliability in real-world applications. Full article

(This article belongs to the Special Issue Control, Optimization and Energy Management in Fuel Cell Hybrid Systems)

15 pages, 274 KiB

Open AccessArticle

Is This the Gate?: J. M. Coetzee’s Elizabeth Costello and Its Operatic Adaptation

by Xingyu Lin

Humanities 2025, 14(3), 55; https://doi.org/10.3390/h14030055 (registering DOI) - 9 Mar 2025

Abstract

Premiered at the 2024 Adelaide Festival, Is This the Gate? is an opera excerpt composed by Nicholas Lens and set to a libretto written by J. M. Coetzee. It is adapted from the last section of Coetzee’s novel Elizabeth Costello (2003), revolving around the eponymous character’s trial before the gate in the afterworld. This article explores the literary, musical and dramaturgical elements of Is This the Gate? and contends that the adaptation, despite its brevity and incompleteness, indexes and reworks some of the most important intertexts, localities and motifs that connect Coetzee’s early and late works. Allusions to Kafka and Dante frame the scenario for Costello in limbo—a state mirroring a writer’s late-in-life predicament—while references to Australia’s weather and fauna reflect Coetzee’s relationship to his South African roots and adopted home. Further, Costello’s conviction that she is “a secretary of the invisible” holds clues to Coetzee’s deployment of voices and fictional personae since his debut, Dusklands (1974). The last few acts of the opera excerpt evoke themes of desire and mortality that chime with Coetzee’s other Costello narratives, including his latest collection, The Pole and Other Stories (2023). The adaptation ends with Costello’s declaration of her subjectivity, which suggests a writer’s yearning and resolution to go beyond the threshold of life and death. Full article

(This article belongs to the Special Issue Music and the Written Word)

20 pages, 5750 KiB

Open AccessArticle

Advanced Insect Detection Network for UAV-Based Biodiversity Monitoring

by Halimjon Khujamatov, Shakhnoza Muksimova, Mirjamol Abdullaev, Jinsoo Cho and Heung-Seok Jeon

Remote Sens. 2025, 17(6), 962; https://doi.org/10.3390/rs17060962 (registering DOI) - 9 Mar 2025

Abstract

The Advanced Insect Detection Network (AIDN), which represents a significant advancement in the application of deep learning for ecological monitoring, is specifically designed to enhance the accuracy and efficiency of insect detection from unmanned aerial vehicle (UAV) imagery. Utilizing a novel architecture that incorporates advanced activation and normalization techniques, multi-scale feature fusion, and a custom-tailored loss function, the AIDN addresses the unique challenges posed by the small size, high mobility, and diverse backgrounds of insects in aerial images. In comprehensive testing against established detection models, the AIDN demonstrated superior performance, achieving 92% precision, 88% recall, an F1-score of 90%, and a mean Average Precision (mAP) score of 89%. These results signify a substantial improvement over traditional models such as YOLO v4, SSD, and Faster R-CNN, which typically show performance metrics approximately 10–15% lower across similar tests. The practical implications of AIDNs are profound, offering significant benefits for agricultural management and biodiversity conservation. By automating the detection and classification processes, the AIDN reduces the labor-intensive tasks of manual insect monitoring, enabling more frequent and accurate data collection. This improvement in data collection quality and frequency enhances decision making in pest management and ecological conservation, leading to more effective interventions and management strategies. The AIDN’s design and capabilities set a new standard in the field, promising scalable and effective solutions for the challenges of UAV-based monitoring. Its ongoing development is expected to integrate additional sensory data and real-time adaptive models to further enhance accuracy and applicability, ensuring its role as a transformative tool in ecological monitoring and environmental science. Full article

►▼ Show Figures

Figure 1

Show export options Show export options

Select all

Export citation of selected articles as:

Error

Oops... you haven't selected anything for export.

Displaying article 1-50 on page 1 of 1628.

Go to page 1 2 3 4 5

Search Results (81,387)

Further Information

Guidelines

MDPI Initiatives

Follow MDPI