PIoU Loss: Towards Accurate Oriented Object Detection in Complex Environments

Chen, Zhiming; Chen, Kean; Lin, Weiyao; See, John; Yu, Hui; Yan, Ke; Yang, Cong

doi:10.1007/978-3-030-58558-7_12

Cited by 213 publications

(75 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…the ResNet, DenseNet, VGG as in [43,44,45]) as backbone instead of the presented convolutional predictor. In the last, iv) when classifying rotated objects, it might be profitable to use a custom loss, such as the Pixels-Intersection-over-Union (PIoU) [46]. The idea is that an oriented bounded box -a one aligned with a rotated object -presents less overlap with the background in complex environments.…”

Section: Resultsmentioning

confidence: 99%

A minimal model for classification of rotated objects with prediction of the angle of rotation

Salas¹,

Dokládal²,

Dokladalova³

2021

Journal of Visual Communication and Image Representation

View full text Add to dashboard Cite

In classification tasks, the robustness against various image transformations remains a crucial property of the Convolutional Neural Networks (CNNs). It can be acquired using the data augmentation. It comes, however, at the price of the risk of overfitting and a considerable increase in training time. Consequently, other ways to endow CNN with invariance to various transformations-and mainly to the rotations-is an intensive field of study. This paper presents a new reduced rotation invariant classification model composed of two parts: a feature representation mapping and a classifier. We provide an insight into the principle and we prove that the proposed model is trainable. This model is smaller in terms of trainable parameters than similar approaches, and has angular prediction capabilities. We illustrate the results on the MNIST and CIFAR-10 datasets. On MNIST, we i) achieve the state of the art of classification on MNIST-rot (with training on MNIST-rot), and ii) improve the results of classification on MNIST-rot (with training on upright MNIST). When trained on CIFAR-10 with upright samples and tested with rotated samples we improve by 20% the state of the art classification results. In all cases, we can predict the rotation angle.

show abstract

Section: Resultsmentioning

confidence: 99%

A minimal model for classification of rotated objects with prediction of the angle of rotation

Salas¹,

Dokládal²,

Dokladalova³

2021

Journal of Visual Communication and Image Representation

View full text Add to dashboard Cite

show abstract

“…A novel AP-Loss [14] is formulated to replace the classification loss, which alleviates foreground-background class imbalance issue. Chen et al [15] propose a novel PIoU Loss to exploit both the angle and IoU for accurate oriented bounding box regression. Confluence [16] selects optimal bounding boxes and removes highly confluent neighboring bounding boxes according to "Manhattan Distance" instead of the conventional IoU.…”

Section: Related Workmentioning

confidence: 99%

SC-RPN: A Strong Correlation Learning Framework for Region Proposal

Zou

Zhang

Peng

et al. 2021

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Current state-of-the-art two-stage detectors heavily rely on region proposals to guide the accurate detection for objects. In previous region proposal approaches, the interaction between different functional modules is correlated weakly, which limits or decreases the performance of region proposal approaches. In this paper, we propose a novel two-stage strong correlation learning framework, abbreviated as SC-RPN, which aims to set up stronger relationship among different modules in the region proposal task. Firstly, we propose a Light-weight IoU-Mask branch to predict intersection-over-union (IoU) mask and refine region classification scores as well, it is used to prevent high-quality region proposals from being filtered. Furthermore, a sampling strategy named Size-Aware Dynamic Sampling (SADS) is proposed to ensure sampling consistency between different stages. In addition, point-based representation is exploited to generate region proposals with stronger fitting ability. Without bells and whistles, SC-RPN achieves AR1000 14.5% higher than that of Region Proposal Network (RPN), surpassing all the existing region proposal approaches. We also integrate SC-RPN into Fast R-CNN and Faster R-CNN to test its effectiveness on object detection task, the experimental results achieve a gain of 3.2% and 3.8% in terms of mAP compared to the original ones.

show abstract

“…O2-DNet [47] proposes a single-stage, anchor-free, and NMS-free model to detect oriented objects by predicting a pair of midlines inside each object. Aiming at the problem of the distance loss function, PIoU [48] uses five parameters of the oriented bounding box to calculate the contribution of each pixel in the area, and then calculates the approximate skew IoU, which greatly improves the detection performance of long objects.…”

Section: B Oriented Object Detectionmentioning

confidence: 99%

Learning Higher Quality Rotation Invariance Features for Multioriented Object Detection in Remote Sensing Images

Zhang

Xiong

Xiao

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Multi-oriented object detection, an important yet challenging task because of the bird's-eye-view perspective, complex background and densely packed objects, is in the spotlight of detection in remote sensing images. Although existing methods have recently experienced substantial progress based on oriented head, they learn little about essential rotation invariance of the objects. In this article, a novel framework is proposed that can learn high-quality rotation invariance features of the multi-oriented objects by three measures. Given a remote sensing image, the MSFF module first merges the global semantic segmentation features predicted by the semantic segmentation branch and the multi-scale features extracted by the backbone with FPN in order to distinguish complex background. Then the discriminative features are used by rotation mainstream, whose structure is similar to Cascade R-CNN and can extract higher-quality rotation invariance features and predict more accurate location information by adaptively adjusting the distribution of the samples through progressive IoU thresholds. And in order to improve the performance of mainstream to predict more accurate oriented bounding box, the horizontal tributaries that can fully leverage the reciprocal relationship between the oriented detection and horizontal detection were added to the latter two stages. Extensive experiments on three public datasets for remote sensing images i.e. Gaofen Airplane, HRSC2016 and DOTA demonstrate that without bells and whistles, the proposed method achieves superior performances compared with the existing state-of-the-art methods for multi-oriented detection. Moreover, our overall system achieves 59.264% mAP of airplane Detection in 2020 Gaofen challenge, ranking 3rd in the final.

show abstract

PIoU Loss: Towards Accurate Oriented Object Detection in Complex Environments

Cited by 213 publications

References 41 publications

A minimal model for classification of rotated objects with prediction of the angle of rotation

A minimal model for classification of rotated objects with prediction of the angle of rotation

SC-RPN: A Strong Correlation Learning Framework for Region Proposal

Learning Higher Quality Rotation Invariance Features for Multioriented Object Detection in Remote Sensing Images

Contact Info

Product

Resources

About