NGIoU Loss: Generalized Intersection over Union Loss Based on a New Bounding Box Regression

Aerial image-based target detection has problems such as low accuracy in multiscale target detection situations, slow detection speed, missed targets and falsely detected targets. To solve this problem, this paper proposes a detection algorithm based on the improved You Only Look Once (YOLO)v3 network architecture from the perspective of model efficiency and applies it to multiscale image-based target detection. First, the K-means clustering algorithm is used to cluster an aerial dataset and optimize the anchor frame parameters of the network to improve the effectiveness of target detection. Second, the feature extraction method of the algorithm is improved, and a feature fusion method is used to establish a multiscale (large-, medium-, and small-scale) prediction layer, which mitigates the problem of small target information loss in deep networks and improves the detection accuracy of the algorithm. Finally, label regularization processing is performed on the predicted value, the generalized intersection over union (GIoU) is used as the bounding box regression loss function, and the focal loss function is integrated into the bounding box confidence loss function, which not only improves the target detection accuracy but also effectively reduces the false detection rate and missed target rate of the algorithm. An experimental comparison on the RSOD and NWPU VHR-10 aerial datasets shows that the detection effect of high-efficiency YOLO (HE-YOLO) is significantly improved compared with that of YOLOv3, and the average detection accuracies are increased by 8.92% and 7.79% on the two datasets, respectively. The algorithm not only shows better detection performance for multiscale targets but also reduces the missed target rate and false detection rate and has good robustness and generalizability.

Intersection over Union (IoU) is the most popular evaluation metric used in the object detection benchmarks. However, there is a gap between optimizing the commonly used distance losses for regressing the parameters of a bounding box and maximizing this metric value. The optimal objective for a metric is the metric itself. In the case of axis-aligned 2D bounding boxes, it can be shown that $IoU$ can be directly used as a regression loss. However, $IoU$ has a plateau making it infeasible to optimize in the case of non-overlapping bounding boxes. In this paper, we address the weaknesses of $IoU$ by introducing a generalized version as both a new loss and a new metric. By incorporating this generalized $IoU$ ($GIoU$) as a loss into the state-of-the art object detection frameworks, we show a consistent improvement on their performance using both the standard, $IoU$ based, and new, $GIoU$ based, performance measures on popular object detection benchmarks such as PASCAL VOC and MS COCO.

The regression loss function is a key factor in the training and optimization process of object detection. The current mainstream regression loss functions are Ln norm loss, IOU loss and CIOU loss. This paper proposes the Scale-Sensitive IOU(SIOU) loss, a new loss function different from the above all, which could solve the issues that the current loss functions cannot distinguish the two bounding boxes in some special cases when the target area scales in one image vary greatly during training process, thereby leading to the improper regression loss calculation and the slowing down of the optimization. An area scale regulating factor Y is added on the basis of CIOU loss to adjust the loss values of the bounding boxes, which could distinguish all the boxes quantitatively in theory thus gets a faster converging speed and better optimization. Through analysis and simulation comparison among the several loss functions, the superiority of SIOU loss is verified. Furthermore, by incorporating SIOU loss into YOLO v4, Faster R-CNN and SSD on the two mainstream aerial remote sensing datasets, i.e., DIOR and NWPU VHR-10, the detection precisions improve by 10.2% than IOU loss and 2.8% than CIOU loss respectively.

An object detector based on convolutional neural network (CNN) has been widely used in the field of computer vision because of its simplicity and efficiency. The average accuracy of CNN model detection results in the object detector is greatly affected by the loss function. The precision of the localization algorithm in the loss function is the main factor affecting the result. Based on the complete intersection over union (CIoU) loss function, an improved penalty function is proposed to improve the localization accuracy. Specifically, the algorithm more comprehensively considers matching bounding boxes between prediction with ground truth, using the proportional relationship of the aspect ratio from both bounding boxes. Under the same aspect ratio of the two bounding boxes, the influence factors of the prediction box on localization accuracy were considered. In this way, the function of the penalty function is strengthened, and localization accuracy of the network model improved. This loss function is called Improved CIoU (ICIoU). Experiments on the Udacity, PASCAL VOC, and MS COCO datasets have demonstrated the effectiveness of ICIoU in improving localization accuracy of network models by using the one-stage object detector YOLOv4. Compared with CIoU, the proposed ICIoU improved average precision (AP) by 0.57% and AP75 by 0.12% on Udacity, AP by 0.26% and AP75 by 1.28% on PASCAL VOC, and AP by 0.06% and AP75 by 0.65% on MS COCO.

Bounding box regression is an important step in the process of object detection, which has a direct impact on the accuracy of model location. The $\ell_{n}$ loss is widely used in bounding box regression, but the criterion of locational accuracy is Intersection over Union (IOU), and there is no strong correlation between them. Although IoU-based loss functions have been proposed successively, such as IoU, GIoU, DIoU and CIoU, which make location loss directly related to evaluation criteria, there are still problems such as slow training convergence speed and inaccurate model location. Based on the IoU loss function, we add normalized corner distance and parameterized width height difference to form a new loss function: LCornerIoU loss. The results of bounding box regression simulation show that LCornerIoU loss is superior to other IoU-based loss functions in terms of convergence speed and positioning accuracy. Experiments on COCO benchmark show that the performance of the Faster R-CNN, RetinaNet, and YOLO-v5s models combined with LCornerIoU loss have consistent performance improvement compared with IoU loss.

As one of the most advanced models in computer vision, the model of the YOLO series has excellent accuracy and efficiency. However, YOLOv5 has achieved great success in object detection. Even since YOLOv5 has a function of image classification, we can still modify and optimize Yolov5 based on model structure, data augmentation, loss function and optimizer, and attention mechanism. Based on the above problems, we replaced the backbone of YOLOv5 with RepVGG as the backbone network of our model. We also used methods such as Cutout, HSV, etc., for image augmentation. We also introduced CBAM (Convolutional Block Attention Module) based on the attention mechanism to improve the model's performance. Besides, we also replaced the GIOU (Generalized Intersection over Union) loss function with the CIOU (Complete-IOU). At the same time, some suitable activation function is selected according to the model, which improves the robustness of the model. Finally, we use the 'Pink Butterfly' dataset to test this model. We were pleasantly surprised to find out about the promising performance. The precision was improved by 4.451 % (from 91.950% to 96.401%). The recall declined 0.413% (from 80.198 % to 79.785%). The [email protected] was improved 2.466 %(from 88.826 % to 91.292%). The [email protected]:0.95 was improved by 2.466 % (from 68.975 % to 73.017 %). Moreover, the model has better robustness.

The traditional welding seam inspection efficiency is low, the model possesses a significant amount of computational parameters, and it is not suitable for the application of small and medium-sized welding seam recognition machines. In view of the wide application and remarkable effect of deep learning in the domain of machine vision, a welding seam defect detection model designed by G-Efficientnet-CA neural network was proposed and optimized. EfficientNet was employed as the core architecture for extracting features to greatly reduce calculation parameters and model volume. The CA attention mechanism is incorporated to enhance the model’s capacity for focusing attention on the weld image, thereby improving its ability to discern and analyze relevant features, and the accuracy is improved. The Generalized Intersection over Union (GIoU) loss function is changed from the original loss function to optimize the calculation of the coincidence degree between the real frame and the predicted frame. The efficient K-Means++ clustering algorithm is utilized to calculate the initial anchor frame which is more suitable for different weld data sets. The experimental outcomes demonstrate that the optimized model, when compared to the YOLOv3 model, exhibits superior effectiveness in detecting the VOC2007 dataset, as evidenced by a notable 12.7% increase in the average precision (mAP) and a significant reduction of 88% in the number of parameters.

The main aim of optical remote sensing image (ORSI) object detection is to ascertain the area as well as class of all targets in given images. Nowadays, object detection methods established from deep learning are progressively applied to the analysis of the information from ORSI. But because of the compound background as well as extensive object scale shown in ORSI, the difficulty of object detection has increased. There are still some challenges to be solved. First, when the two bounding boxes don't overlap, current indexes like intersection over union (IoU) cannot identify the distance between them. Secondly, the existing methods using convolutional neural network (CNN) could not take advantage of the characteristics of multi-level features so that these methods are not very good at recognizing small-sized objects. To deal with these above problems, this paper comes up with a ORSI object detection method established on multi-level feature fusion as well as improved boundary box regression compared to IoU. Firstly, a new measure called generalized intersection over union (GIoU) is practiced. That metric could measure the similarity betwixt two bounding boxes, regardless of whether they overlap or not. At the same time, we also directly use GIoU as loss. Lastly, a multi-level feature fusion structure is practiced and it is also integrated into the existing convolutional neural network. In this way, our method could take advantage of the multi-level features. Quantitative comparison of the addressed method with the baseline approach on large-scale dataset named Dior are conducted. The comparison with the most advanced method shows that our method has reached the most advanced performance.

A crucial role is played by steering-angle prediction in the control of autonomous vehicles (AVs). It mainly includes the prediction and control of the steering angle. However, the prediction accuracy and calculation efficiency of traditional YOLOv5 are limited. For the control of the steering angle, angular velocity is difficult to measure, and the angle control effect is affected by external disturbances and unknown friction. This paper proposes a lightweight steering angle prediction network model called YOLOv5Ms, based on YOLOv5, aiming to achieve accurate prediction while enhancing computational efficiency. Additionally, an adaptive output feedback control scheme with output constraints based on neural networks is proposed to regulate the predicted steering angle using the YOLOv5Ms algorithm effectively. Firstly, given that most lane-line data sets consist of simulated images and lack diversity, a novel lane data set derived from real roads is manually created to train the proposed network model. To improve real-time accuracy in steering-angle prediction and enhance effectiveness in steering control, we update the bounding box regression loss function with the generalized intersection over union (GIoU) to Shape-IoU_Loss as a better-converging regression loss function for bounding-box improvement. The YOLOv5Ms model achieves a 30.34% reduction in weight storage space while simultaneously improving accuracy by 7.38% compared to the YOLOv5s model. Furthermore, an adaptive output feedback control scheme with output constraints based on neural networks is introduced to regulate the predicted steering angle via YOLOv5Ms effectively. Moreover, utilizing the backstepping control method and introducing the Lyapunov barrier function enables us to design an adaptive neural network output feedback controller with output constraints. Finally, a strict stability analysis based on Lyapunov stability theory ensures the boundedness of all signals within the closed-loop system. Numerical simulations and experiments have shown that the proposed method provides a 39.16% better root mean squared error (RMSE) score than traditional backstepping control, and it achieves good estimation performance for angles, angular velocity, and unknown disturbances.

Object detection in degraded environments such as low-light and nighttime conditions remains a challenging task, as conventional computer vision techniques often fail to achieve high precision and robust performance. With the increasing adoption of deep learning, this paper aims to enhance object detection under such adverse conditions by proposing an improved version of YOLOv11, named DRF-YOLO (Degradation-Robust and Feature-enhanced YOLO). The proposed framework incorporates three innovative components: (1) a lightweight Cross Stage Partial Multi-Scale Edge Enhancement (CSP-MSEE) module that combines multi-scale feature extraction with edge enhancement to strengthen feature representation; (2) a Focal Modulation attention mechanism that improves the network’s responsiveness to target regions and contextual information; and (3) a self-developed Dynamic Interaction Head (DIH) that enhances detection accuracy and spatial adaptability for small objects. In addition, a lightweight unsupervised image enhancement algorithm, Zero-DCE (Zero-Reference Deep Curve Estimation), is introduced prior to training to improve image contrast and detail, and Generalized Intersection over Union (GIoU) is employed as the bounding box regression loss. To evaluate the effectiveness of DRF-YOLO, experiments are conducted on two representative low-light datasets: ExDark and the nighttime subset of BDD100K, which include images of vehicles, pedestrians, and other road objects. Results show that DRF-YOLO achieves improvements of 3.4% and 2.3% in mAP@0.5 compared with the original YOLOv11, demonstrating enhanced robustness and accuracy in degraded environments while maintaining lightweight efficiency.

An lightweight anchor-free synthetic aperture radar (SAR) ship detector is proposed in this paper. Firstly, According to the SAR image characteristics, ResSARNet with only 0.69M parameters was proposed. Then, with ResSARNet as the backbone network, four improvements have been made to fully convolutional one-stage object detection (FCOS): Center-ness on bounding box regression branch, center sampling, Generalized Intersection over Union (GIoU) loss and adaptive training sample selection (ATSS). To verify the effectiveness of the proposed method, sufficient experiments have been conducted on the widely used SAR Ship Detection Dataset (SSDD). The results show that, the proposed network can achieve 61.5% average precision (AP) and 70.9% average recall (AR) with only 1.17M parameters, which are only 6.5% and 3.2% for RetinaNet@ResNet-18 and RetinaNet@ResNet-50, respectively.

In object detection algorithms, the bounding box regression (BBR) loss directly influences the model's accuracy in predicting object positions. Initially, we conducted simulation experiments on the proposed intersection over union-based loss. Through in-depth analysis, we discovered that the width and height components of the distance term could hinder BBR. To address this, we proposed the AIoU-v1 loss, which decouples the width and height components of the distance term, thereby preventing the suppression of BBR loss by the distance term. Additionally, issues such as the imbalance between sample quantity and sample quality in the dataset, as well as labeling errors, can adversely affect BBR. To tackle these dataset problems, we designed an adaptive dynamic non-monotonic focusing mechanism with strong robustness and wide applicability. Finally, we proposed a post-processing algorithm that combines fusion and non-maximum suppression, resulting in more accurate bounding boxes during the post-processing stage. Our source code and data are available at https://github.com/Wenhua-Zheng/AIOU.

In order to improve the performance of the target detection model based on convolutional neural network, and make the bounding box representing the object location contain as complete original target information as possible, this paper improved the loss function of the bounding box regression part of the YOLOv3 network. A new metric, Asymmetric Intersection over Union (AIoU), is proposed to measure the similarity between the prediction box and the truth. Theoretical analysis shows that AIoU based loss function can better guide the direction of bounding box regression after introducing asymmetry. Experiments show that the improved YOLOv3 network achieves higher accuracy and maintains fast convergence in human target detection. The examples of bounding box regression on the test dataset show that the target information loss can be reduced by the proposed method. In addition, distinguishing between the prediction and the truth in calculation may have reference significance for the loss function design of network models for other tasks

Object detection is a fundamental task in computer vision. Bounding box regression (BBR) losses are critical to detector performance. However, evaluation measures that rely on the Intersection over Union (IoU) between the predicted and ground truth boxes are highly sensitive to positional deviations, which can hinder optimization. To alleviate this issue, we propose ATDIoU, a novel arctangent-differential loss for bounding-box regression. ATDIoU computes distance similarity between a predicted and a ground truth box by modeling the distances between their corresponding vertices as a two-dimensional arctangent differential distribution (ATD). This arctangent differential-based design mitigates bounding box drift and reduces sensitivity to localization errors. As a result, it guides the model to learn target positions more effectively. We evaluate ATDIoU by integrating it into YOLOv6 and conducting experiments on PASCAL VOC and VisDrone2019. The results demonstrate that ATDIoU yields improvements of 1.4% and 0.7% in mean average precision (mAP) relative to MPDIoU.

It is challenging to accurately detect flexible objects with arbitrary orientation from monitoring images in power grid maintenance and inspection sites. This is because these images exhibit a significant imbalance between the foreground and background, which can lead to low detection accuracy when using a horizontal bounding box (HBB) as the detector in general object detection algorithms. Existing multi-oriented detection algorithms that use irregular polygons as the detector can improve accuracy to some extent, but their accuracy is limited due to boundary problems during the training process. This paper proposes a rotation-adaptive YOLOv5 (R_YOLOv5) with a rotated bounding box (RBB) to detect flexible objects with arbitrary orientation, effectively addressing the above issues and achieving high accuracy. Firstly, a long-side representation method is used to add the degree of freedom (DOF) for bounding boxes, enabling accurate detection of flexible objects with large spans, deformable shapes, and small foreground-to-background ratios. Furthermore, the further boundary problem induced by the proposed bounding box strategy is overcome by using classification discretization and symmetric function mapping methods. Finally, the loss function is optimized to ensure training convergence for the new bounding box. To meet various practical requirements, we propose four models with different scales based on YOLOv5, namely R_YOLOv5s, R_YOLOv5m, R_YOLOv5l, and R_YOLOv5x. Experimental results demonstrate that these four models achieve mean average precision (mAP) values of 0.712, 0.731, 0.736, and 0.745 on the DOTA-v1.5 dataset and 0.579, 0.629, 0.689, and 0.713 on our self-built FO dataset, exhibiting higher recognition accuracy and a stronger generalization ability. Among them, R_YOLOv5x achieves a mAP that is about 6.84% higher than ReDet on the DOTAv-1.5 dataset and at least 2% higher than the original YOLOv5 model on the FO dataset.