VarifocalNet: An IoU-aware Dense Object Detector

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.

A prototype of a visual-to-auditory sensory substitution device based on a video camera with two lenses was prepared. Software to convert the signal from a video camera into an audio signal in mono and stereo modes was developed.To assess the developed system, an experimental study with 30 blindfolded sighted participants was conducted. 15 persons were tested in mono mode, 15 - in stereo mode. All persons were trained to use the visual-to-auditory sensory substitution system. During the experiment, participants were to locate a white plastic cube with dimensions of 4×4×4 cm3 on a working surface. The researcher placed the cube in one of 20 positions on the working surface in a pseudo-random order. To assess the accuracy of the cube localization, deviations along the X- and Y-axes and absolute deviations were calculated. The general dynamics of localization accuracy was positive both in mono and stereo modes. Absolute deviation and X-axis deviation were significantly higher in stereo mode; there was no significant difference in Y-axis deviation between modes. On average, participants tended to underestimate the distance to the cube when it was on the left, right, or far side of the working surface, and overestimate the distance to the cube when it was on the near side of the working surface. Tests demonstrated that the accuracy of object localization in stereo mode can be improved by increasing the time for training the participants and by showing them more presentations. The results of the study can be used to develop assistive techniques for people with visual impairments, to manufacture medical equipment, and create brain-computer interfaces.

Object detection and tracking is one of the key applications of wireless sensor networks (WSNs). The key issues associated with this application include network lifetime, object detection and localization accuracy. To ensure the high quality of the service, there should be a trade-off between energy efficiency and detection accuracy, which is challenging in a resource-constrained WSN. Most researchers have enhanced the application lifetime while achieving target detection accuracy at the cost of high node density. They neither considered the system cost nor the object localization accuracy. Some researchers focused on object detection accuracy while achieving energy efficiency by limiting the detection to a predefined target trajectory. In particular, some researchers only focused on node clustering and node scheduling for energy efficiency. In this study, we proposed a mobile object detection and tracking framework named the Energy Efficient Object Detection and Tracking Framework (EEODTF) for heterogeneous WSNs, which minimizes energy consumption during tracking while not affecting the object detection and localization accuracy. It focuses on achieving energy efficiency via node optimization, mobile node trajectory optimization, node clustering, data reporting optimization and detection optimization. We compared the performance of the EEODTF with the Energy Efficient Tracking and Localization of Object (EETLO) model and the Particle-Swarm-Optimization-based Energy Efficient Target Tracking Model (PSOEETTM). It was found that the EEODTF is more energy efficient than the EETLO and PSOEETTM models.

Object detectors are conventionally trained by a weighted sum of classification and localization losses. Recent studies (e.g., predicting IoU with an auxiliary head, Generalized Focal Loss, Rank & Sort Loss) have shown that forcing these two loss terms to interact with each other in non-conventional ways creates a useful inductive bias and improves performance. Inspired by these works, we focus on the correlation between classification and localization and make two main contributions: (i) We provide an analysis about the effects of correlation between classification and localization tasks in object detectors. We identify why correlation affects the performance of various NMS-based and NMS-free detectors, and we devise measures to evaluate the effect of correlation and use them to analyze common detectors. (ii) Motivated by our observations, e.g., that NMS-free detectors can also benefit from correlation, we propose Correlation Loss, a novel plug-in loss function that improves the performance of various object detectors by directly optimizing correlation coefficients: E.g., Correlation Loss on Sparse R-CNN, an NMS-free method, yields 1.6 AP gain on COCO and 1.8 AP gain on Cityscapes dataset. Our best model on Sparse R-CNN reaches 51.0 AP without test-time augmentation on COCO test-dev, reaching state-of-the-art. Code is available at: https://github.com/fehmikahraman/CorrLoss.

Object detection in aerial images is an active yet challenging task in computer vision because of the bird’s-eye view perspective, the highly complex backgrounds, and the variant appearances of objects. Especially when detecting densely packed objects in aerial images, methods relying on horizontal proposals for common object detection often introduce mismatches between the Region of Interests (RoIs) and objects. This leads to the common misalignment between the final object classification confidence and localization accuracy. In this paper, we propose a RoI Transformer to address these problems. The core idea of RoI Transformer is to apply spatial transformations on RoIs and learn the transformation parameters under the supervision of oriented bounding box (OBB) annotations. RoI Transformer is with lightweight and can be easily embedded into detectors for oriented object detection. Simply apply the RoI Transformer to light head RCNN has achieved state-of-the-art performances on two common and challenging aerial datasets, i.e., DOTA and HRSC2016, with a neglectable reduction to detection speed. Our RoI Transformer exceeds the deformable Position Sensitive RoI pooling when oriented bounding-box annotations are available. Extensive experiments have also validated the flexibility and effectiveness of our RoI Transformer.

Weakly supervised object localization (WSOL) aims to localize objects with only image-level labels. Previous methods often try to utilize feature maps and classification weights to localize objects using image level annotations indirectly. In this paper, we demonstrate that weakly supervised object localization should be divided into two parts: class-agnostic object localization and object classification. For class-agnostic object localization, we should use class-agnostic methods to generate noisy pseudo annotations and then perform bounding box regression on them without class labels. We propose the pseudo supervised object localization (PSOL) method as a new way to solve WSOL. Our PSOL models have good transferability across different datasets without fine-tuning. With generated pseudo bounding boxes, we achieve 58.00% localization accuracy on ImageNet and 74.97% localization accuracy on CUB-200, which have a large edge over previous models.

Deep learning have been playing an increasing important role in ship detection in synthetic aperture radar (SAR) image because of its high accuracy, fast speed and no need of manual design features. It is difficult to set appropriate anchors when using anchor-based method because of multi-resolution SAR images and multi-scale ships, thus resulting in degradation of detection performance in operational applications. Recently, a lot of anchor-free architecture have been put forward, among which most of them apply horizontal bounding boxes (HBBs) to encode ship location, leading to inaccurate location description and decreased detection performance especially for arbitrary-oriented ships in complex scenes. Therefore, an anchor-free detector for arbitrary-oriented ship detection in SAR images, called R-FCOS is proposed in this paper. Specifically, based on the FCOS baseline method, ship aspect angle and rotate regression branch are introduced to regress oriented bounding boxes (OBBs) for arbitrary-oriented ships detection. Besides, to alleviate the effects of feature misalignment problem, ratio-ness branch is proposed to predict the aspect ratio of ship, which is used to filter and refine the OBBs in the post-processing step to improve detection accuracy. Finally, we carry out experiments on the SSDD+ data set to verify the effectiveness of our method. It turns out that our method obtains mAP 90.65% under an IoU threshold of 0.5.

Loss functions, such as the IoU Loss function and the GIoU (Generalized Intersection over Union) Loss function have been put forward to replace regression loss functions commonly used in regression loss calculation. GIoU Loss alleviates the vanishing gradient in the case of the non-overlapping, but it will completely degenerate into the IoU Loss function when bounding boxes overlap totally, which fails to achieve the optimization effect. To solve this problem, some improvements are proposed in this paper on the basis of the GIoU Loss function, taking into account the overlap rate of complete overlap of bounding boxes. In PASCAL VOC data, the experimental results demonstrate that the AP of NGIoU Loss function in the YOLOv4 model is 47.68%, 1.15% higher than that of the GIoU Loss function, and the highest map value is 86.79% in the YOLOv5 model.

The labeling cost of large number of bounding boxes is one of the main challenges for training modern object detectors. To reduce the dependence on expensive bounding box annotations, we propose a new semi-supervised object detection formulation, in which a few seed box level annotations and a large scale of image level annotations are used to train the detector. We adopt a training-mining framework, which is widely used in weakly supervised object detection tasks. However, the mining process inherently introduces various kinds of labelling noises: false negatives, false positives and inaccurate boundaries, which can be harmful for training the standard object detectors (e.g. Faster RCNN). We propose a novel NOise Tolerant Ensemble RCNN (NOTE-RCNN) object detector to handle such noisy labels. Comparing to standard Faster RCNN, it contains three highlights: an ensemble of two classification heads and a distillation head to avoid overfitting on noisy labels and improve the mining precision, masking the negative sample loss in box predictor to avoid the harm of false negative labels, and training box regression head only on seed annotations to eliminate the harm from inaccurate boundaries of mined bounding boxes. We evaluate the methods on ILSVRC 2013 and MSCOCO 2017 dataset; we observe that the detection accuracy consistently improves as we iterate between mining and training steps, and state-of-the-art performance is achieved.

The key to a successful cascade architecture for precise instance segmentation is to fully leverage the relationship between bounding box detection and mask segmentation across multiple stages. Although modern instance segmentation cascades achieve leading performance, they mainly make use of a unidirectional relationship, i.e., mask segmentation can benefit from iteratively refined bounding box detection. In this paper, we investigate an alternative direction, i.e., how to take the advantage of precise mask segmentation for bounding box detection in a cascade architecture. We propose a Deeply Shape-guided Cascade (DSC) for instance segmentation, which iteratively imposes the shape guidances extracted from mask prediction at the previous stage on bounding box detection at current stage. It forms a bi-directional relationship between the two tasks by introducing three key components: (1) Initial shape guidance: A mask-supervised Region Proposal Network (mPRN) with the ability to generate class-agnostic masks; (2) Explicit shape guidance: A mask-guided region-of-interest (RoI) feature extractor, which employs mask segmentation at previous stage to focus feature extraction at current stage within a region aligned well with the shape of the instance-of-interest rather than a rectangular RoI; (3) Implicit shape guidance: A feature fusion operation which feeds intermediate mask features at previous stage to the bounding box head at current stage. Experimental results show that DSC outperforms the state-of-the-art instance segmentation cascade, Hybrid Task Cascade (HTC), by a large margin and achieves 51.8 box AP and 45.5 mask AP on COCO test-dev. The code is released at: https://github.com/hding2455/DSC.

One-stage object detectors are simple and efficient; however, they cannot extract sufficient object features due to simplistic structures. At the same time, the classification score cannot reflect the actual positioning of the candidate box. Therefore, it is not accurate to use classification score only as the candidate box position score in non-maximum suppression (NMS) stage. These two shortcomings degrade the detection accuracy. In this paper, a novel feature pyramid architecture named refined feature pyramid network (ReFPN) is introduced to obtain better object features. ReFPN designs a refined module which is parallel with feature pyramid network (FPN) to extract the semantic features of objects, and then the extraction of features are used to optimize the features of FPN by summation. In addition, we design the refined center-ness (RCenter-ness) branch that predicts the position score of each point on the feature map to improve the localization accuracy. The predicted position score is multiplied by the classification score to obtain the final position score that has a stronger correlation with localization accuracy. The final position score is inputted to the subsequent NMS, which improves localization accuracy. The proposed method in this paper is named ReFPN-FCOS. The sufficient experiments on COCO2017 datasets demonstrate the effectiveness of ReFPN-FCOS on improving classification accuracy and localization accuracy. The average precisions of this method achieve 1.1% and 1.3 % higher than those of FCOS, when using ResNet50 and ResNet101 as backbone respectively. Code download link: https://github.com/xjl-le/mmdete.

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.

Automatic detection of weapons is significant for improving security and well being of individuals, nonetheless, it is a difficult task due to large variety of size, shape and appearance of weapons. View point variations and occlusion also are reasons which makes this task more difficult. Further, the current object detection algorithms process rectangular areas, however a slender and long rifle may really cover just a little portion of area and the rest may contain unessential details. To overcome these problem, we propose a CNN architecture for Orientation Aware Weapons Detection, which provides oriented bounding box with improved weapons detection performance. The proposed model provides orientation not only using angle as classification problem by dividing angle into eight classes but also angle as regression problem. For training our model for weapon detection a new dataset comprising of total 7801 weapons images is gathered from the web and then manually annotated with position oriented bounding boxes. Our dataset provides not only oriented bounding box as ground truth but also horizontal bounding box. We also provide our dataset in multiple formats of modern object detectors for further research in this area. The proposed model is evaluated on this dataset, and the comparative analysis with off-the shelf object detectors yields superior performance of proposed model, measured with standard evaluation strategies. The dataset and the model implementation are made publicly available at this link: https://bit.ly/2TyZICF .

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.

This paper proposes an approach for object class localization which goes beyond bounding boxes, as it also determines the outline of the object. Unlike most current localization methods, our approach does not require any hypothesis parameter space to be defined. Instead, it directly generates, evaluates and clusters shape masks. Thus, the presented framework produces more informative results for object class localization. For example, it easily learns and detects possible object viewpoints and articulations, which are often well characterized by the object outline. We evaluate the proposed approach on the challenging natural-scene Graz-02 object classes dataset. The results demonstrate the extended localization capabilities of our method.