YOLO-V2 (You Only Look Once)

The paper presents a pragmatic approach towards developing and analysing a GUI-based system performance towards detecting and counting tracked multi-type vehicles in mixed traffic conditions using the improved you only look once v3 (YOLOv3) model. It addresses the issue related to accurately localising smaller vehicles in an occluded scenario and environmental conditions. In the training phase of the proposed work stochastic gradient descent (SGD) optimises the network. Further, to detect small vehicles four different scales feature mapping are performed to extract more fine-grained features for accurate detection, followed by concatenating upsampled layers with lower layers to improve the detection performance of low-resolution images. The intersection over union (IoU) approach detects every vehicle in the subsequent frame by assigning a unique ID to classify detected vehicles into five specific classes bus, truck, car, motorbike, and bicycle. Further, the SORT algorithm tracks and counts the detected vehicles. Experimental result on the common objects in context (COCO) dataset shows an improvement in the mean average precision (mAP) by 11% compared to the existing YOLOv3 technique.

Object detection is a fundamental problem in computer vision, with applications spanning self-driving cars, surveillance systems, medical imaging, robotics, and smart cities. Among the myriad of algorithms developed for this task, the You Only Look Once (YOLO) family stands out for its ability to perform real-time and accurate object detection. This article provides a comprehensive analysis of the YOLO algorithm series, from YOLOv1 to YOLOv8, evaluating them across key performance metrics, including precision, recall, mean Average Precision (mAP), frames per second (FPS), and overall effectiveness. Unlike traditional two-stage detectors such as R-CNN, YOLO formulates object detection as a single regression problem: a single pass over the image simultaneously predicts bounding boxes and class probabilities. This end-to-end design enables YOLO to achieve high speed while maintaining a competitive accuracy-efficiency trade-off. We examine architectural innovations across YOLO versions, including batch normalization, anchor boxes, residual blocks, feature pyramid networks, and attention mechanisms, and discuss their impact on performance. Lightweight models (e.g., YOLOv5-Nano, YOLOv8-Small) are explored with a focus on their suitability for mobile and embedded systems, highlighting YOLO’s adaptability to resource-constrained environments. Challenges such as small object detection, occlusion, and domain-specific tuning are also addressed. This article serves as a practical guide for researchers, developers, and practitioners aiming to leverage YOLO for real-world object detection tasks.

In this paper, we are interested in the object detection problem with high accuracy and fast prediction. To achieve this, we have modified the existing object detection algorithm, we design a new feature extraction network, predict detection target based on multiple feature maps and improve the anchor box. Particularly, since DenseNet can combine the output of previous convolutional layers by using skip connection operation to make prediction more accurate, we introduce DenseNet to improve the YOLOv3-Tiny network. Furthermore, we add a prediction layer to the improved network to detect more true targets. Finally, we clustered the ground truth of specific dataset to get the anchor box for training to improve the detection results of the improved network. We demonstrate the effectiveness of our network by using Pascal VOC datasets. The experimental results show that the improved network increases the detection accuracy by 14% and the detection speed is 63 frames per second in the video detection. The detection accuracy is improved to a higher level and the detection speed still meeting the real-time detection requirement, which shows that the proposed method is robust and effective. Anchor box, DenseNet, Video detection

Automatic target recognition (ATR) is a challenging task for several computer vision applications. It requires efficient, accurate, and robust methods for target detection and target identification. Deep learning has shown great success in many computer vision applications involving color RGB images. However, the performance of these networks in ATR with infrared sensor data needs further investigation. In this paper, we propose a multistage automatic target detection and recognition (ATDR) system that performs both target detection and target classification on infrared (IR) imagery using deep learning. Our system processes large IR image frames where targets take <1 % of the total number of pixels. First, we train a state-of-the-art object detector you only look once (YOLO) to localize all potential targets in the input image frame. Then, we train a convolutional neural network (CNN) to identify these detections as targets or false alarms. In this second phase, we adapt and analyze the performance of three CNN architectures: a compact and fully connected CNN, VGG16 with batch normalization, and a wide residual neural network (WRN). We also explore the use of a loss function that optimizes directly the area under the receiver operating characteristic (ROC) curve (AUC), and adapt it to our ATR application. To enhance the robustness of the proposed ATR to perturbation and variations introduced during the detection stage, we train our CNN classifiers on automatically detected targets using YOLO, in addition to ground truth bounding boxes and apply selected data augmentation techniques. To simulate real testing environments, where the spatial location of the targets within the image frame is unknown, only YOLO-detected boxes are used during validation. We evaluate our ATDR on a real benchmark dataset that includes different vehicles captured at different resolutions. Our experiments have shown that YOLO can detect most of the targets at the expense of generating a high number of false alarms. We show that the VGG-16 network with batch normalization, which is the best performing model, can correctly identify the classes of the targets, as well as classify the majority of YOLO’s false detections into an additional nontarget class. We also show that the proposed training modification to optimize an AUC-based loss function for ATR proved to be advantageous mainly in identifying difficult targets.

The construction of a smart autonomous mobile robot system that can operate in an anonymous environment can be difficult. Building an autonomous mobile robot requires collecting data from sensors and cameras in order to recognize the surrounding environment, get knowledge of existing landmarks, and figure out its position and orientation. Many research studies have been carried out in recent years providing algorithms for achieving autonomous motion of mobile robot systems and object detection. The most common platform utilized for robot motion design is ROS (Robot Operating System) and for object detection YOLO (You Only Look Once) is a common choice. Additionally, the autonomous system needs motion planning in a systematic way to avoid obstacles by using LIDAR or other sensing methods generating the feedback signals required for control of motion around occupied spaces. One of the most effective and versatile feedback systems for mobile robot control can be constructed using visual feedback, which relies on a high-resolution camera that processes the image in real-time and the result is fed back to the control system for decision making. Visual feedback can be employed alone without sensors for feedback control of autonomous mobile robots, and it can be used for multiple tasks like path planning, obstacle avoidance, object detection, object recognition in addition to motion control. In this research, results of design and construction of an autonomous mobile rescue robot for finding victims through video data are presented. It is aimed to find victims via applying video as a data collection and feedback generation tool in the autonomous mobile robot control system. The visual feedback and YOLO v3 are used not only for finding victims but also for motion control. Our autonomous mobile robot is designed to move smoothly to the victim's direction and avoids obstacles by using ROS Global Planner. The robot is tested in multiple scenarios in different locations, various obstacle and victim numbers, and satisfactory results with good performances are obtained.

This paper presents a new approach by training and improving a convolutional neural network (CNN) based on You Only Look Once version 2 (YOLOv2) to efficiently detect fuel trucks from images in embedded systems. The proposed method considers the entire image area for strong object detection compared with existing methods that only focus on the image area where the class object exists to predict its probability to be in a class. The loss function for CNN is improved to enhance effective learning, especially when only a limited amount of data is available for training. The class probability can be learned by improving the loss function although the anchor boxes are not in the center of the target object. The learning process of the model can be in a limited range and achieve rapid convergence although the sizes of the initial anchor and target boundary boxes are different. Experimental results of various fuel truck images show the efficiency of the proposed approach under different detection scenarios of real fuel trucks. The detection rate of the proposed method is approximately 4% higher than the YOLOv2 object detection method. The proposed method is suitable to monitor long country borders using unmanned drones.

CasRPN: Cascade region proposal network for visual tracking

In this chapter, a human detection system based on unsupervised learning method K-means clustering followed by deep learning approach You Only Look Once (YOLO) on thermal imagery has been proposed. Generally, images in the visible spectrum are used to conduct such human detection, which are not suitable for nighttime due to low visibility, hence for evaluation of our system. Hence, long wave infrared (LWIR) images have been used to implement the proposed work in this chapter. The system follows a two-step approach of generating anchor boxes using K-means clustering and then using those anchor boxes in 252 layered single shot detector (YOLO) to predict proper boundary boxes. The dataset of such images is provided by FLIR company. The dataset contains 6822 images for training purposes and 757 images for the validation. This proposed system can be used for real-time object detection as YOLO can achieve much higher rate of processing when compared to traditional method like HAAR cascade classifier in long wave infrared imagery (LWIR).

Software was developed in Python for the digital processing of object images before their detection using the You Only Look Once (YOLO) artificial neural network. Detection was carried out on object images, specifically vehicles and other road users. The preprocessing of images was performed through histogram equalization and local contrast enhancement. The original images were read from video cameras or graphic files. Image detection was performed using the medium-sized YOLO convolutional neural network model, YOLOv8m. As a result of detection, the studied objects, particularly cars, were highlighted with rectangular bounding boxes. To evaluate the accuracy of object detection, several metrics (parameters) were used: Precision, Recall, Intersection over Union (IoU), F1, and Average Precision (AP). The YOLO network was pre-trained on the Common Objects in Context dataset. The YOLO neural network returns the object recognition confidence, the coordinates of its center (x, y) on the image, and its width and height (w, h). The detected objects are marked with rectangular bounding boxes. The positioning of objects in the images was manually performed using the Roboflow tool, which allows for accurate spatial positioning of each object. The study of the impact of image preprocessing on the accuracy of object detection by the YOLO network was conducted using a typical traffic scene image. After histogram equalization, the histogram becomes more symmetrical and uniform, improving the visual quality of both dark and bright areas. This led to improved object detection accuracy across almost all metrics. The increase in detection accuracy by the IoU parameter is 0.169. The summary improvement in detection accuracy by the IoU, F1, AP parameters is 0.502. It was shown that after image preprocessing, detection accuracy increased, and all vehicles were correctly detected. The developed software can be used in various computer vision systems as well as in Internet of Things systems for remote monitoring and control.

This paper proposes a Contextual Action Recognition System for classroom monitoring that integrates spatial and temporal modeling with contextual analysis for enhanced robustness and interpretability. YOLO (You Only Look Once) is used for spatial feature extraction, while ConvLSTM and Temporal Convolutional Network (TCN) capture temporal patterns. An attention-based fusion mechanism adaptively weights their outputs. A hybrid proximity detection approach, combining overlap assessment with Intersection over Union (IoU) thresholding outperforms traditional IoU-based methods, especially in perspectivedistorted scenes. The model achieved strong results: precision of 0.90, recall of 0.87, F 1 -score of 0.86, and $\mathrm{mAP} \text{@} 0.5$ of 0.92. It outperformed baselines like SlowFast and ACRN on datasets including AVA, UCF101-24, and JHMDB-21, demonstrating superior spatial relationship detection.

Solar Radio Bursts (SRBs) are generally observed in dynamic spectra and have five major spectral classes, labelled Type I to Type V depending on their shape and extent in frequency and time. Due to their complex characterization, a challenge in solar radio physics is the automatic detection and classification of such radio bursts. Classification of SRBs has become fundamental in recent years due to large data rates generated by advanced radio telescopes such as the LOw-Frequency ARray, (LOFAR). Current state-of-the-art algorithms implement the Hough or Radon transform as a means of detecting predefined parametric shapes in images. These algorithms achieve up to 84% accuracy, depending on the Type of radio burst being classified. Other techniques include procedures that rely on Constant-False-Alarm-Rate detection, which is essentially detection of radio bursts using a de-noising and adaptive threshold in dynamic spectra. It works well for a variety of different Types of radio bursts and achieves an accuracy of up to 70%. In this research, we are introducing a Convolutional Neural Network (CNN) named You Only Look Once v2 (YOLOv2) for solar radio burst classification. By using Type III simulation methods we can train the algorithm to classify real Type III solar radio bursts in real-time at an accuracy of 82.63% with a maximum 77 frames per second (fps).

Lesion segmentation in dermoscopic images significantly enhances the diagnostic performance of AI-based classification models. However, conventional methods often require pixel-level annotations, which are resource-intensive and prone to errors caused by external artifacts, such as hair and skin markings. We propose a hybrid framework called SAM-enhanced YOLO, which integrates the Segment Anything Model (SAM) with You Only Look Once (YOLO) for precise pixel-level segmentation. This method combines YOLO's efficient lesion localization with SAM's advanced zero-shot segmentation capabilities. To further validate the framework, we compared it against traditional methods, including GrabCut and Otsu's thresholding, as well as SAM used without YOLO (SAM-only). For SAM-only, lesion segmentation was initialized at the image center to simulate a typical dermoscopic imaging setup. SAM-enhanced YOLO demonstrated superior segmentation performance, achieving an Intersection over Union (IoU) of 0.738 and an F1-score (the harmonic mean of precision and recall) of 0.833, compared to 0.578 and 0.683 with SAM-only, respectively. This represents a 28% improvement in IoU and a 22% improvement in F1-score compared to SAM-only. The results were consistent across lesion shapes and contrast conditions, with SAM-enhanced YOLO exhibiting the lowest variability and highest robustness among the evaluated methods. By reducing the need for pixel-level annotations and outperforming both standalone SAM and traditional methods, SAM-enhanced YOLO provides a scalable and resource-efficient solution for dermoscopic lesion segmentation. This framework holds significant potential for improving diagnostic workflows in clinical and resource-limited settings.

Object detection for blind zones is critical to ensuring the driving safety of heavy trucks. We propose a scheme to realize object detection in the blind zones of heavy trucks based on the improved you-only-look-once (YOLO) v3 network. First, according to the actual detection requirements, the targets are determined to establish a new data set of persons, cars, and fallen pedestrians, with a focus on small and medium objects. Subsequently, the network structure is optimized, and the features are enhanced by combining the shallow and deep convolution information of the Darknet platform. In this way, the feature propagation can be effectively enhanced, feature reuse can be promoted, and the network performance for small object detection can be improved. Furthermore, new anchors are obtained by clustering the data set using the K-means technique to improve the accuracy of the detection frame positioning. In the test stage, detection is performed using the trained model. The test results demonstrate that the proposed improved YOLO v3 network is superior to the original YOLO v3 model in terms of the blind zone detection and can satisfy the accuracy and real-time requirements with an accuracy of 94% and runtime of 13.792 ms / frame. Moreover, the mean average precision value for the improved model is 87.82%, which is 2.79% higher than that of the original YOLO v3 network.

Tiny objects in remote sensing images only have a few pixels, and the detection difficulty is much higher than that of regular objects. General object detectors lack effective extraction of tiny object features, and are sensitive to the Intersection-over-Union (IoU) calculation and the threshold setting in the prediction stage. Therefore, it is particularly important to design a tiny-object-specific detector that can avoid the above problems. This article proposes the network JSDNet by learning the geometric Jensen-Shannon (JS) divergence representation between Gaussian distributions. First, the Swin Transformer model is integrated into the feature extraction stage as the backbone to improve the feature extraction capability of JSDNet for tiny objects. Second, the anchor box and ground-truth are modeled as two two-dimensional (2D) Gaussian distributions, so that the tiny object is represented as a statistical distribution model. Then, in view of the sensitivity problem faced by the IoU calculation for tiny objects, the JSDM module is designed as a regression sub-network, and the geometric JS divergence between two Gaussian distributions is derived from the perspective of information geometry to guide the regression prediction of anchor boxes. Experiments on the AI-TOD and DOTA datasets show that JSDNet can achieve superior detection performance for tiny objects compared to state-of-the-art general object detectors.

This study was aimed to explore the adoption value of magnetic resonance imaging (MRI) under convolutional neural networks (CNN) in the therapeutic effect of nasopharyngeal carcinoma (NPC) radiotherapy. A total of 54 NPC patients were recruited. CNN was employed to perform 3D visualization processing on magnetic resonance (MR) images of NPC patients. MRI changes were analyzed before and after the patient received radiotherapy. The image segmentation and radiotherapy effects of CNN were evaluated by the Recall, intersection over union (IOU), postoperative apparent diffusion coefficient (ADC), and diagnostic coincidence rate. Moreover, gradient vector flow (GVF) algorithm, fuzzy c-means (FCM), and SegNet were adopted for comparative evaluation. Recall of CNN was 94.89% and the IOU was 84.16%, which was remarkably different from other algorithms (P &lt; 0.05). After analysis of the MRI images of patients receiving radiotherapy, ADC of local residual patients was 1.108 ± 0.097 measured by CNN, the ADC was 1.826 ± 0.115, and the missed diagnosis rate was only 7.14%. In summary, CNN had a good effect on the localization and segmentation of NPC patients, and can accurately evaluate the effect of patients receiving radiotherapy, which can assist clinical diagnosis and treatment of NPC.