Search

Wildfires pose significant environmental and societal threats, necessitating improved early detection methods. This study investigates the effectiveness of integrating real-time object detection deep learning models (YOLOv8 and RT-DETR) with advanced data augmentation techniques, including StyleGAN2-ADA, for wildfire smoke detection. We evaluated model performance on datasets enhanced with fundamental transformations and synthetic images, focusing on detection accuracy. YOLOv8X demonstrated superior overall performance with AP@0.33 of 0.962 and AP@0.5 of 0.900, while RT-DETR-X excelled in small object detection with a 0.983 detection rate. Data augmentation, particularly StyleGAN2-ADA, significantly enhanced model performance across various metrics. Our approach reduced average detection times to 1.52 min for YOLOv8X and 2.40 min for RT-DETR-X, outperforming previous methods. The models demonstrated robust performance under challenging conditions, like fog and camera noise, providing reassurance of their effectiveness. While false positives remain a challenge, these advancements contribute significantly to early wildfire smoke detection capabilities, potentially mitigating wildfire impacts through faster response times. This research establishes a foundation for more effective wildfire management strategies and underscores the potential of deep learning applications in environmental monitoring. [ABSTRACT FROM AUTHOR]

Copyright of Baghdad Science Journal is the property of Republic of Iraq Ministry of Higher Education & Scientific Research (MOHESR) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Harbin University of Science & Technology is the property of Journal of Harbin University of Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Deep learning models have significantly improved object detection, which is essential for visual sensing. However, their increasing complexity results in higher latency and resource consumption, making real-time object detection challenging. In order to address the challenge, we propose a new lightweight filtering method called L-filter to predict empty video frames that include no object of interest (e.g., vehicles) with high accuracy via hybrid time series analysis. L-filter drops those frames deemed empty and conducts object detection for nonempty frames only, significantly enhancing the frame processing rate and scalability of real-time object detection. Our evaluation demonstrates that L-filter improves the frame processing rate by 31–47% for a single traffic video stream compared to three standalone state-of-the-art object detection models without L-filter. Additionally, L-filter significantly enhances scalability; it can process up to six concurrent video streams in one commodity GPU, supporting over 57 fps per stream, by working alongside the fastest object detection model among the three models. [ABSTRACT FROM AUTHOR]

In automotive adhesive defect detection, manual inspection suffers from low efficiency and blind spots in human vision, which affects the performance of parts. Therefore, automated detection methods are particularly important. To address the issue of adhesive defects significantly impacting production during automated gluing processes, we propose an adhesive defect detection method for automotive applications based on the improved YOLOv8 (named YOLOv8n-SSE). First, we used the SSE (skip squeeze and excitation) attention mechanism in the backbone part to dynamically adjust the importance of different channels in our model and allow our model to selectively focus on important features. Then, the original bounding box loss function is replaced by the WIoU loss function. Experimental results demonstrate that this method improves the mAP50 of the original YOLOv8n by 3.25% and achieves an average detection speed of 7.9ms per image, equivalent to 126.58 frames per second (FPS), meeting the real-time defect detection requirements. [ABSTRACT FROM AUTHOR]

Deep learning technology for real-time small object detection in aerial images can be used in various industrial environments such as real-time traffic surveillance and military reconnaissance. However, detecting small objects with few pixels and low resolution remains a challenging problem that requires performance improvement. To improve the performance of small object detection, we propose DCEF   2 -YOLO. Our proposed method enables efficient real-time small object detection by using a deformable convolution (DFConv) module and an efficient feature fusion structure to maximize the use of the internal feature information of objects. DFConv preserves small object information by preventing the mixing of object information with the background. The optimized feature fusion structure produces high-quality feature maps for efficient real-time small object detection while maximizing the use of limited information. Additionally, modifying the input data processing stage and reducing the detection layer to suit small object detection also contributes to performance improvement. When compared to the performance of the latest YOLO-based models (such as DCN-YOLO and YOLOv7), DCEF   2 -YOLO outperforms them, with a mAP of +6.1% on the DOTA-v1.0 test set, +0.3% on the NWPU VHR-10 test set, and +1.5% on the VEDAI512 test set. Furthermore, it has a fast processing speed of 120.48 FPS with an RTX3090 for 512 × 512 images, making it suitable for real-time small object detection tasks. [ABSTRACT FROM AUTHOR]

Real-time object detection and tracking is an active area of aerial remote sensing research that enables many environmental and ecological monitoring and preservation applications. Despite the development of several solutions tailored for these specific applications, trade-offs between cost efficiency and feature richness persist. This paper proposes a lightweight, low-cost, and modular approach to real-time object detection and instance tracking, enabling a wide gamut of use cases. By integrating real-time object detection models with affordable embedded hardware, we present a system that uses image metadata to perform geolocation on detected objects, enabling real-time applications due to minimal computational overhead. This algorithm generates cleaner ’areas of interest’ based on geolocated detections filtered by a clustering algorithm to remove false positives. In our findings, this proved a viable solution with real-time processing speeds and GPS positioning accuracy within a meter. While there is room for improvement, our proposed pipeline represents a significant step forward in lowering the costs involved with applying computer vision to conservation applications.

With a long-term usage, highways usually suffer from diverse pavement defects, which causes burdensome pavement defect detections on vast areas. To address this issue, the real-time and vehicle-mounted technique has been proposed and proved to be an efficient solution. However, as the detection systems are deployed on edge devices in complex environments, there exist several practical challenges in terms of real-time speed, detection performance and computational overhead. This study first provides a dataset captured by real-time and vehicle-mounted camera system. The height and angle where the high frame rate camera locates has been carefully designed to achieve the acquisitions of real-time images on highways. Then, this study proposes a novel lightweight network model named Real-time Pavement Detection of You Only Look Once (RPD-YOLO) with a lightweight Light C3 Ghost (LCG) block and an LCG Path Aggregation Network (LCG-PAN) neck structure, which can fully reduce the computational overheads and maintain a high precision and high speed during detection. Through a series of comparison experiments with current models, RPD-YOLO excels in overall balanced performance and can be deployed in resource-constrained devices to achieve real-time pavement defect detection.

Traditional palm oil production methods for evaluating fruit bunches (FFBs) are inefficient, costly, and have limited coverage. This study evaluates the performance of various YOLO models and other state-of-the-art object detection models using a novel dataset of oil palm fresh fruit bunches in plantations, captured in the plantation regions of Central Kalimantan Province, Indonesia. The dataset includes five ripeness classes (abnormal, ripe, underripe, unripe, and flower) and presents challenges such as partially visible objects, low contrast scenes, occluded and small objects, and blurry images. The proposed YOLOv8s Depthwise model was compared with other YOLO models, including YOLOv6s, YOLOv6l, YOLOv7 Tiny, YOLOv7l, YOLOv8s, and YOLOv8l. YOLOv8s Depthwise demonstrated a balanced performance, with a compact size (10.6 MB), fast inference time (0.027 seconds), and strong detection accuracy (mAP50 at 0.75, mAP50-95 at 0.481). Its rapid convergence and low training loss highlighted its efficiency, completing training in the shortest time of 2 hours, 18 minutes, and 30 seconds. Furthermore, it achieved low Mean Absolute Error (MAE) of 0.164 and Root Mean Square Error (RMSE) of 0.4, indicating precise counting capability. Hyperparameter tuning revealed that the YOLOv8s Depthwise model achieved optimal performance using the SGD optimizer with a batch size of 16 and a learning rate of 0.001, showing the best balance between accuracy and training efficiency. Data augmentation positively impacted model performance, resulting in improved performance metrics across various models. When evaluated against other state-of-the-art models on the same dataset, including Faster RCNN, SSD MobileNetV2, YOLOv4, and YOLOv9, YOLOv8s Depthwise surpassed other state-of-the-art models, including Faster R-CNN, SSD MobileNetV2, YOLOv4, and EfficientDet-D0 from previous research, in terms of speed, accuracy, and efficiency, making it ideal for real-time palm oil harvesting applications.

Vehicle object detection using UAV images is a crucial undertaking in urban traffic management and the advancement of autonomous driving technologies. Conventional networks fail to achieve accuary in detecting vehicle objects from a drone’s perspective due to the significant variations in the size of the target items, unequal distribution of their positions in the image, and image degradation induced by the drone’s movement.In order to surmount this challenge, this research suggests an enhanced TOOD object detection model named Drone-TOOD. The first proposal is to create a lightweight network skeleton called CSPRegNet by merging CSPblock with Regblock. Secondly, we incorporate Regblock into CSPPAFPN to enhance CSPRegPAFPN and incorporate EVCblock at the upsampling location of deep features to capture corner area details and minimize the degradation of feature information. In addition, a efficient task decomposition attention module is also proposed to enhance the interaction ability of positioning and classification tasks. This task decomposition module can highlight the characteristics of a specific task while retaining the characteristics of another task, thereby improving detection capabilities. Experiments conducted on the Drone Vision Challenge Benchmark (VisDrone) demonstrate that the enhanced model can obtain superior performance compared to TOOD. The average precision (mAP) achieved by our approach is 64%, surpassing TOOD by 7.9%. The frames per second (FPS) stayed constant at 27.2. Drone-TOOD demonstrates superior performance compared to other lightweight models on the VisDrone-2021 dataset. In order to demonstrate the robustness of our approach, we additionally performed ablation experiments and conducted tests on the UAV Detection and Tracking Dataset (UAVDT), resulting in an achieved mean average precision (mAP) of 64.6%. Furthermore, Drone-TOOD possesses a total of parameter approximately 21.7 M.

Autonomous vehicles are becoming increasingly popular in a variety of applications, including waste collection and fire detection. In this work, we present the design and implementation of an autonomous vehicle for these tasks in urban environments. The vehicle is equipped with sensors and control algorithms to navigate, detect and collect plastic bottle wastes, and detect fires in real-time. The system uses an off-the-shelf, small-sized, battery-operated vehicle, a simple conveyor belt, and a vision-based, computerized system. Machine learning (ML-) based vision tasks are implemented to direct the vehicle to waste locations and initiate the waste removal process. A fire detection and alarm system are also incorporated, using a camera and machine learning algorithms to detect flames automatically. The vehicle was tested in a simulated urban environment, and the results demonstrate its effectiveness in waste material collection and fire detection. The proposed system has the potential to improve the efficiency and safety of such tasks in urban areas.

Wildfires pose significant environmental and societal threats, necessitating improved early detection methods. This study investigates the effectiveness of integrating real-time object detection deep learning models (YOLOv8 and RT-DETR) with advanced data augmentation techniques, including StyleGAN2-ADA, for wildfire smoke detection. We evaluated model performance on datasets enhanced with fundamental transformations and synthetic images, focusing on detection accuracy. YOLOv8X demonstrated superior overall performance with AP@0.33 of 0.962 and AP@0.5 of 0.900, while RT-DETR-X excelled in small object detection with a 0.983 detection rate. Data augmentation, particularly StyleGAN2-ADA, significantly enhanced model performance across various metrics. Our approach reduced average detection times to 1.52 min for YOLOv8X and 2.40 min for RT-DETR-X, outperforming previous methods. The models demonstrated robust performance under challenging conditions, like fog and camera noise, providing reassurance of their effectiveness. While false positives remain a challenge, these advancements contribute significantly to early wildfire smoke detection capabilities, potentially mitigating wildfire impacts through faster response times. This research establishes a foundation for more effective wildfire management strategies and underscores the potential of deep learning applications in environmental monitoring.

Advances in information and signal processing, driven by artificial intelligence techniques and recent breakthroughs in deep learning, have significantly impacted autonomous driving by enhancing safety and reducing the dependence on human intervention. Generally, prevailing ADASs (advanced driver assistance systems) incorporate costly components, making them financially unattainable for a substantial portion of the population. This paper proposes a solution: an embedded system designed for real-time pedestrian and priority sign detection, offering affordability and universal applicability across various vehicles. The suggested system, which comprises two cameras, an NVIDIA Jetson Nano B01 low-power edge device and an LCD (liquid crystal system) display, ensures seamless integration into a vehicle without occupying substantial space and provides a cost-effective alternative. The primary focus of this research is addressing accidents caused by the failure to yield priority to other drivers or pedestrians. Our study stands out from existing research by concurrently addressing traffic sign recognition and pedestrian detection, concentrating on identifying five crucial objects: pedestrians, pedestrian crossings (signs and road paintings separately), stop signs, and give way signs. Object detection was executed using a lightweight, custom-trained CNN (convolutional neural network) known as SSD (Single Shot Detector)-MobileNet, implemented on the Jetson Nano. To tailor the model for this specific application, the pre-trained neural network underwent training on our custom dataset consisting of images captured on the road under diverse lighting and traffic conditions. The outcomes of the proposed system offer promising results, positioning it as a viable candidate for real-time implementation; its contributions are noteworthy in advancing the safety and accessibility of autonomous driving technologies. [ABSTRACT FROM AUTHOR]

In the field of unmanned systems, the combination of artificial intelligence with self-operating functionalities is becoming increasingly important. This study introduces a new method for autonomously detecting humans in indoor environments using unmanned aerial vehicles, utilizing the advanced techniques of a deep learning framework commonly known as "You Only Look Once" (YOLO). The key contribution of this research is the development of a new model (YOLO-IHD), specifically designed for human detection in indoor using drones. This model is created using a unique dataset gathered from aerial vehicle footage in various indoor environments. It significantly improves the accuracy of detecting people in these complex environments. The model achieves a notable advancement in autonomous monitoring and search-and-rescue operations, highlighting its importance for tasks that require precise human detection. The improved performance of the new model is due to its optimized convolutional layers and an attention mechanism that process complex visual data from indoor environments. This results in more dependable operation in critical situations like disaster response and indoor rescue missions. Moreover, when combined with an accelerating processing library, the model shows enhanced real-time detection capabilities and operates effectively in a real-world environment with a custom designed indoor drone. This research lays the groundwork for future enhancements designed to significantly increase the model's accuracy and the reliability of indoor human detection in real-time drone applications. [ABSTRACT FROM AUTHOR]

Nowadays, individuals are experiencing an era characterized by a pandemic situation. Given the challenging circumstances face these days, it is imperative to develop a real-time model that automatically detects whether candidates are wearing ornaments or carrying any devices before appearing for an entrance examination, as it serves as a vital precautionary measure for our health and safety. The proposed system will function in two stages: firstly, it will employ the YOLOv7 algorithm, which operates on Convolutional Neural Networks (CNN), to detect any gadgets or ornaments worn by the candidate. Additionally, we incorporated Mobile Net and HairNet into the code, utilizing TensorFlow's Keras API. Notably, all the datasets used in this system have been created from scratch. Upon successfully detecting gadgets such as wristwatches, Bluetooth devices, earrings, and other materials, the system will promptly display a warning message, denying entry to the respective individuals. Moreover, the proposed system includes real-time monitoring of the examination hall through CCTV cameras. As a cost-effective, time-saving, and accurate solution, it can be implemented across various examinations to ensure candidates' and invigilators' safety and adhere to examination norms. [ABSTRACT FROM AUTHOR]

Object tracking, an important technology in the field of image processing and computer vision, is used to continuously track a specific object or person in an image. This technology may be effective in identifying the same person within one image, but it has limitations in handling multiple images owing to the difficulty in identifying whether the object appearing inother images is the same.Whentracking the same object using two ormore images, there must be a way to determine that objects existing in different images are the same object. Therefore, this paper attempts to determine the same object present in different images using color information among the unique information of the object. Thus, this study proposes a multiple-object-tracking method using histogram stamp extraction in closed-circuit television applications. The proposed method determines the presence or absence of a target object in an image by comparing the similarity between the image containing the target object and other images. To this end, a unique color value of the target object is extracted based on its color distribution in the image using three methods: mean, mode, and interquartile range. The Top-N accuracy method is used to analyze the accuracy of each method, and the results show that the mean method had an accuracy of 93.5% (Top-2). Furthermore, the positive prediction value experimental results show that the accuracy of the mean method was 65.7%. As a result of the analysis, it is possible to detect and track the same object present in different images using the unique color of the object. Through the results, it is possible to track the same object that can minimize manpower without using personal information when detecting objects in different images. In the last response speed experiment, it was shown that when the mean was used, the color extraction of the object was possible in real time with 0.016954 s. Through this, it is possible to detect and track the same object in real time when using the proposed method. [ABSTRACT FROM AUTHOR]

Home service robots operating indoors, such as inside houses and offices, require the real-time and accurate identification and location of target objects to perform service tasks efficiently. However, images captured by visual sensors while in motion states usually contain varying degrees of blurriness, presenting a significant challenge for object detection. In particular, daily life scenes contain small objects like fruits and tableware, which are often occluded, further complicating object recognition and positioning. A dynamic and real-time object detection algorithm is proposed for home service robots. This is composed of an image deblurring algorithm and an object detection algorithm. To improve the clarity of motion-blurred images, the DA-Multi-DCGAN algorithm is proposed. It comprises an embedded dynamic adjustment mechanism and a multimodal multiscale fusion structure based on robot motion and surrounding environmental information, enabling the deblurring processing of images that are captured under different motion states. Compared with DeblurGAN, DA-Multi-DCGAN had a 5.07 improvement in Peak Signal-to-Noise Ratio (PSNR) and a 0.022 improvement in Structural Similarity (SSIM). An AT-LI-YOLO method is proposed for small and occluded object detection. Based on depthwise separable convolution, this method highlights key areas and integrates salient features by embedding the attention module in the AT-Resblock to improve the sensitivity and detection precision of small objects and partially occluded objects. It also employs a lightweight network unit Lightblock to reduce the network's parameters and computational complexity, which improves its computational efficiency. Compared with YOLOv3, the mean average precision (mAP) of AT-LI-YOLO increased by 3.19%, and the detection precision of small objects, such as apples and oranges and partially occluded objects, increased by 19.12% and 29.52%, respectively. Moreover, the model inference efficiency had a 7 ms reduction in processing time. Based on the typical home activities of older people and children, the dataset Grasp-17 was established for the training and testing of the proposed method. Using the TensorRT neural network inference engine of the developed service robot prototype, the proposed dynamic and real-time object detection algorithm required 29 ms, which meets the real-time requirement of smooth vision. [ABSTRACT FROM AUTHOR]

Convolutional Neural Networks (CNNs) are well-studied and commonly used for the problem of object detection thanks to their increased accuracy. However, high accuracy on its own says little about the effective performance of CNN-based models, especially when real-time detection tasks are involved. To the best of our knowledge, there has not been sufficient evaluation of the available methods in terms of their speed/accuracy trade-off. This work performs a review and hands-on evaluation of the most fundamental object detection models on the Common Objects in Context (COCO) dataset with respect to this trade-off, their memory footprint, and computational and storage costs. In addition, we review available datasets for medical mask detection and train YOLOv5 on the Properly Wearing Masked Faces Dataset (PWMFD). Next, we test and evaluate a set of specific optimization techniques, transfer learning, data augmentations, and attention mechanisms, and we report on their effect for real-time mask detection. Based on our findings, we propose an optimized model based on YOLOv5s using transfer learning for the detection of correctly and incorrectly worn medical masks that surpassed more than two times in speed (69 frames per second) the state-of-the-art model SE-YOLOv3 on the PWMFD while maintaining the same level of mean Average Precision (67%).

Video object detection plays a pivotal role in various applications, from surveillance to autonomous vehicles. This research addresses the need for real-time object detection in videos using advanced deep learning models. The current landscape of object detection techniques often struggles to maintain efficiency in processing video streams, leading to delays and resource-intensive computations. This study aims to bridge this gap by proposing a novel methodology for real-time object detection in videos. With the surge in video data across domains, the demand for swift and accurate object detection in real-time has become imperative. Existing methods face challenges in balancing speed and precision, prompting the exploration of more robust solutions. This research endeavors to enhance the efficiency of video object detection, offering a timely and accurate approach to address contemporary demands. The primary challenge lies in achieving real-time object detection without compromising accuracy. Traditional methods often compromise speed for precision, leading to inadequate performance in dynamic video environments. This study seeks to overcome this dilemma by introducing a methodology that optimizes both speed and accuracy, catering to the real-time constraints of video processing. Despite the advancements in object detection, a notable research gap exists in the domain of real-time video object detection. Existing models exhibit limitations in adapting to the dynamic nature of video streams, necessitating the development of novel methodologies. This research aims to fill this void by proposing an innovative approach that addresses the specific challenges posed by real-time video data. The proposed methodology integrates state-of-the-art deep learning models, optimizing them for real-time video object detection. Leveraging advanced architectures and streamlining the inference process, the model aims to provide accurate detections at unparalleled speeds. Additionally, a novel data augmentation technique is introduced to enhance the model's adaptability to dynamic video scenarios. Preliminary results demonstrate the effectiveness of the proposed methodology, showcasing a significant improvement in both real-time processing speed and object detection accuracy. The model exhibits promising performance across diverse video datasets, highlighting its potential to outperform existing methods in real-world applications. [ABSTRACT FROM AUTHOR]

Convolutional Neural Networks (CNNs) are well-studied and commonly used for the problem of object detection thanks to their increased accuracy. However, high accuracy on its own says little about the effective performance of CNN-based models, especially when real-time detection tasks are involved. To the best of our knowledge, there has not been sufficient evaluation of the available methods in terms of their speed/accuracy trade-off. This work performs a review and hands-on evaluation of the most fundamental object detection models on the Common Objects in Context (COCO) dataset with respect to this trade-off, their memory footprint, and computational and storage costs. In addition, we review available datasets for medical mask detection and train YOLOv5 on the Properly Wearing Masked Faces Dataset (PWMFD). Next, we test and evaluate a set of specific optimization techniques, transfer learning, data augmentations, and attention mechanisms, and we report on their effect for real-time mask detection. Based on our findings, we propose an optimized model based on YOLOv5s using transfer learning for the detection of correctly and incorrectly worn medical masks that surpassed more than two times in speed (69 frames per second) the state-of-the-art model SE-YOLOv3 on the PWMFD while maintaining the same level of mean Average Precision (67%). [ABSTRACT FROM AUTHOR]

Deep learning models have significantly improved object detection, which is essential for visual sensing. However, their increasing complexity results in higher latency and resource consumption, making real-time object detection challenging. In order to address the challenge, we propose a new lightweight filtering method called L-filter to predict empty video frames that include no object of interest (e.g., vehicles) with high accuracy via hybrid time series analysis. L-filter drops those frames deemed empty and conducts object detection for nonempty frames only, significantly enhancing the frame processing rate and scalability of real-time object detection. Our evaluation demonstrates that L-filter improves the frame processing rate by 31–47% for a single traffic video stream compared to three standalone state-of-the-art object detection models without L-filter. Additionally, L-filter significantly enhances scalability; it can process up to six concurrent video streams in one commodity GPU, supporting over 57 fps per stream, by working alongside the fastest object detection model among the three models.

Deep learning technology for real-time small object detection in aerial images can be used in various industrial environments such as real-time traffic surveillance and military reconnaissance. However, detecting small objects with few pixels and low resolution remains a challenging problem that requires performance improvement. To improve the performance of small object detection, we propose DCEF 2-YOLO. Our proposed method enables efficient real-time small object detection by using a deformable convolution (DFConv) module and an efficient feature fusion structure to maximize the use of the internal feature information of objects. DFConv preserves small object information by preventing the mixing of object information with the background. The optimized feature fusion structure produces high-quality feature maps for efficient real-time small object detection while maximizing the use of limited information. Additionally, modifying the input data processing stage and reducing the detection layer to suit small object detection also contributes to performance improvement. When compared to the performance of the latest YOLO-based models (such as DCN-YOLO and YOLOv7), DCEF 2-YOLO outperforms them, with a mAP of +6.1% on the DOTA-v1.0 test set, +0.3% on the NWPU VHR-10 test set, and +1.5% on the VEDAI512 test set. Furthermore, it has a fast processing speed of 120.48 FPS with an RTX3090 for 512 × 512 images, making it suitable for real-time small object detection tasks.

With the recent announcement of the Defense Innovation 4.0 Basic Plan by the Ministry of National Defense, the role and operational scope of drones are expanding as a key force in AI-based, unmanned, and autonomous systems. Consequently, the significance of real-time object detection technology is emphasized as drones take on diverse missions, including delivering, analyzing, and assessing real-time, target-related information. The emergence of recent deep learning has led to substantial advancements in the field of computer vision, particularly in object detection. Deep learning-based object detection is actively being researched, with a focus on algorithms suited for embedded and mobile environments such as drones. This research predominantly aims to develop deep learning-based object detection models that ensure real-time performance and accurately identify objects’ various forms and sizes. Recent object-detection models have been categorized into backbone networks, neck networks, and head networks. By utilizing these three network components, design considerations can be tailored to fulfill the requirements of drone operations. In this paper, we investigate the technology trends of deep learning network models that can be loaded into drones for real-time object detection. Thus we contribute to strengthening effective drone operation in military operations and supporting research and decision-making processes.

The accuracy of current deep learning algorithms has certainly increased. However, deploying deep learning networks on edge devices with limited resources is challenging due to their inherent depth and high parameter count. Here, we proposed an improved YOLO model based on an attention mechanism and receptive field (RFA-YOLO) model, applying the MobileNeXt network as the backbone to reduce parameters and complexity, adopting the Receptive Field Block (RFB) and Efficient Channel Attention (ECA) modules to improve the detection accuracy of multi-scale and small objects. Meanwhile, an FPGA-based model deployment solution was proposed to implement parallel acceleration and low-power deployment of the detection algorithm model, which achieved real-time object detection for optical remote sensing images. We implement the proposed DPU and Vitis AI-based object detection algorithms with FPGA deployment to achieve low power consumption and real-time performance requirements. Experimental results on DIOR dataset demonstrate the effectiveness and superiority of our RFA-YOLO model for object detection algorithms. Moreover, to evaluate the performance of the proposed hardware implementation, it was implemented on a Xilinx ZCU104 board. Results of the experiments for hardware and software simulation show that our DPU-based hardware implementation are more power efficient than central processing units (CPUs) and graphics processing units (GPUs), and have the potential to be applied to onboard processing systems with limited resources and power consumption. [ABSTRACT FROM AUTHOR]

Artificial intelligence (AI) is shaping manufacturing to make it smarter, intelligent, and autonomous. Presently, flexible robots have been introduced that collaborate with humans on the shop floor to enhance productivity and efficiency. Object classification and pose estimation in an autonomous robotic system are crucial problems for proper grasping. Extensive research is being conducted to achieve low-cost, computationally efficient, and real-time assessments. However, most of the existing approaches are computationally expensive and constrained to previous knowledge of the 3D structure of an object. This article presents an AI-based solution, which generalizes cuboid- and cylindrical-shaped objects' grasping in real-time, irrespective of the dimensions. The AI algorithm has achieved an average precision of 89.44% and 82.43% for cuboid- and cylindrical-shaped objects. It is identified without the knowledge of the objects' 3D model. The pose is estimated in real-time, accurately. The integrated solution has been implemented in a robotic system fitted with two grippers, a conveyor system, and sensors. Results of several experiments have been reported in this article, which validates the solution. The proposed methodology has achieved 100% accuracy during our experiments to grasp objects on the conveyor belt. [ABSTRACT FROM AUTHOR]

One of the tasks of robotics is to develop a robot's ability to perform specific actions for as long as possible without human assistance. One such step is to open different types of doors. This task is essential for any operation that involves moving a robot from one room to another. This paper proposes a versatile and computationally efficient algorithm for an autonomous mobile robot opening different types of doors, using machine learning methods. The latter include the YOLOv5 object detection model, the RANSAC iterative method for estimating the mathematical model parameters, and the DBSCAN clustering algorithm. Alternative clustering methods are also compared. The proposed algorithm was explored and tested in simulation and on a real robot manufactured by SOMATIC version Dalek. The percentage of successful doors opened out of the total number of attempts was used as an accuracy metric. The proposed algorithm reached an accuracy of 95% in 100 attempts. The result of testing the door-handle detection algorithm on simulated data was an error of 1.98 mm in 10,000 samples. That is, the average distance from the door handle found by the detector to the real one was 1.98 mm. The proposed algorithm has shown high accuracy and the ability to be applied in real time for opening different types of doors. [ABSTRACT FROM AUTHOR]

Recent developments in drone technology have led to the widespread use of unmanned aerial vehicles (UAVs). In particular, UAVs are often used in reconnaissance to detect objects such as missing persons in large areas. However, traditional systems use only one UAV to search for missing persons in a large area. In addition, object detection is performed after flight or manually because detection requires high computing power. In this paper, a reconnaissance drone system using multiple UAVs is proposed. The proposed multi-UAV reconnaissance system performs real-time object detection on each UAV. The real-time object detection results from each UAV are received by the ground control system (GCS) to stitch the images. To enable real-time object detection in individual UAVs, the filter pruning method is applied to the YOLOv5 model, and the model uses 40% fewer parameters than the existing baseline model. The lightweight YOLOv5 model achieves approximately 11.73 FPS on the Jetson Xaiver NX using a mission computer. Moreover, the proposed image stitching method enables image stitching by effectively matching features using additional information generated by UAVs. The UAV flight tests show that the proposed reconnaissance system can monitor and detect objects in real time over large areas.

Transformers have become increasingly prevalent in computer vision research, especially for object detection. To accurately and efficiently distinguish the stem end of pomelo from its black spots, we propose a hierarchical feature detector, which reconfigures the self-attention model, with high detection accuracy. We designed the combination attention module and the hierarchical feature fusion module that utilize multi-scale features to improve detection performance. We created a dataset in COCO format and annotated two types of detection targets: the stem end and the black spot. Experimental results on our pomelo dataset confirm that HFD's results are comparable to those of state-of-the-art one-stage detectors such as YOLO v4 and YOLO v5 and transformer-based detectors such as DETR, Deformable DETR, and YOLOS. It achieves 89.65% mAP at 70.92 FPS with 100.34 M parameters. [ABSTRACT FROM AUTHOR]

Object detection is one of the most popular areas today. The new models of object detection are created continuously and applied in various fields that help to modernize the old solutions in practice. In this manuscript, the focus has been on investigating the influence of training parameters on similar object detection: image resolution, batch size, iteration number, and color of images. The results of the model have been applied in real-time object detection using mobile devices. The new construction detail dataset has been collected and used in experimental investigation. The models have been evaluated by two measures: the accuracy of each prepared model has been measured; results of real-time object detection on testing data, where the recognition ratio has been calculated. The highest influence on the accuracy of the created models has the iteration number chosen in the training process and the resolution of the images. The higher the resolution of the images that have been selected, the lower the accuracy that has been obtained. The small iteration number leads to the model not being well trained and the accuracy of the models being very low. Slightly better results were obtained when the color images were used. [ABSTRACT FROM AUTHOR]

In the field of unmanned systems, the combination of artificial intelligence with self-operating functionalities is becoming increasingly important. This study introduces a new method for autonomously detecting humans in indoor environments using unmanned aerial vehicles, utilizing the advanced techniques of a deep learning framework commonly known as “You Only Look Once” (YOLO). The key contribution of this research is the development of a new model (YOLO-IHD), specifically designed for human detection in indoor using drones. This model is created using a unique dataset gathered from aerial vehicle footage in various indoor environments. It significantly improves the accuracy of detecting people in these complex environments. The model achieves a notable advancement in autonomous monitoring and search-and-rescue operations, highlighting its importance for tasks that require precise human detection. The improved performance of the new model is due to its optimized convolutional layers and an attention mechanism that process complex visual data from indoor environments. This results in more dependable operation in critical situations like disaster response and indoor rescue missions. Moreover, when combined with an accelerating processing library, the model shows enhanced real-time detection capabilities and operates effectively in a real-world environment with a custom designed indoor drone. This research lays the groundwork for future enhancements designed to significantly increase the model’s accuracy and the reliability of indoor human detection in real-time drone applications.

Abstract The decrease in the possibilities of increasing the arable agricultural areas in the world and the continuous increase in the population have led those who are engaged in plant production to seek ways to make maximum use of the existing agricultural areas. One of these ways is mixed sowing systems. It is very difficult to sow species with different grain sizes in mixtures. Special sowing machines are needed for this aim. Because of this reason, the article aims to be a guide for artificial intelligence capable of mixed sowing in forage crops. In the research, it is found that there are some differences between YOLOv4-tiny and Y0L0v4 models as Precision, Recall, F1-score, TP, FP, FN scores. For the YOLOv4-tiny model, these scores were found as 0.99, 1.00, 0.99, 90, 1, 0, respectively and the scores for the Y0L0v4 model were 1.00, 1.00, 1.00, 90, 0, 0. According to the YOLOv4-tiny and YOLOv4 tests in the lab, suggesting that the YOLOv4-tiny is faster, and the YOLOv4 is more reliable in terms of all these factors combined. This research establishes a standard for real-time recognition of forage crops based on current technology at NVIDIA Jetson TX2 due to its high performance and low power consumption and a high-performance computer with CUDA support.