Arbitrary-Oriented Object Detection with Circular Smooth Label

Yang, Xue; Yan, Junchi

doi:10.1007/978-3-030-58598-3_40

Cited by 429 publications

(294 citation statements)

References 50 publications

Supporting

Mentioning

293

Contrasting

Order By: Relevance

“…[13] formally identify the rotation sensitivity error in rotational object detectors and devise a special treatment to ensure the loss continuity and regression consistency. [14] make a thorough exploration of boundary problem exits in regression method and propose a circular smooth label (CSL) technique to handle it, while SCRDet [10] add an IoU constant factor to the smooth L1 loss to address boundary problem. Densely Coded Labels (DCL) [36] futher explores classification methodology and pushes its frontier work [14] by making detectors sensitive to angular distance and object's aspect ratio.…”

Section: Regression Problems In Angle-based Methodsmentioning

confidence: 99%

“…Therefore, RetinaNet is very suitable to do extension as a basic framework. There are many rotation detectors are built on RetinaNet [9], [14], [18], [36]. Specifically, [9] proposed Rotation RetinaNet to represent oriented rectangle (x, y, w, h, θ), which is a five-parameter system and calls for predicting an additional angular offset in the regression subnet.…”

Section: Rotation Detector Built On Retinanetmentioning

confidence: 99%

“…Specifically, [9] proposed Rotation RetinaNet to represent oriented rectangle (x, y, w, h, θ), which is a five-parameter system and calls for predicting an additional angular offset in the regression subnet. [14] adding an extra classification network at the box regression subnetwork of RetinaNet for more accurate angular predicting; [18] designed a matching compensation factor to focal loss defined in RetinaNet, which helps to improve the localization performance. RetinaNet based detectors helps to realize real-time inference, which is essential for many remote scenarios.…”

Section: Rotation Detector Built On Retinanetmentioning

confidence: 99%

“…Through carefully study of the above works, we find that regression problems in the angle-based method is mainly caused by two reasons, as we already summarized in Fig. 1: i) angular periodicity [10], [14] brings negative consequences, when ideal predictions are beyond the defined range and redundancy of ideal anchors; ii) regression inconsistent [13] cased by different measurement between angle and other parameter and misassignment of weight and height when they exchanged. This paper summarizes the above two issuess as regression uncertainty problem.…”

Section: E Comparison and Summarymentioning

confidence: 99%

“…cerned regression-based methods, the most commonly used technique is to rotate the bounding box by introducing an extra angle prediction. However from a comprehensive point of view, these angle-based methods remain challenging due to two fundamental issues in practice: i) angular periodicity which leads to boundary case [10], [13], [14] and redundancy of anchor. As a typical example in Fig.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Bounding Box Projection for Regression Uncertainty in Oriented Object Detection

Xiang

Tang³

et al. 2021

IEEE Access

View full text Add to dashboard Cite

Oriented object detection has recently attracted increasing attention for its importance in aerial image processing. Popular detection methods for oriented and densely packed objects usually utilize the rotation angle to reduce the overlap of bounding boxes over the horizontal line. However, those anglebased methods remain challenging due to the angular periodicity and regression inconsistent, which are uniformly summarized as regression uncertainty. Previous methods usually heal the regression uncertainty problem by adding complex limits on the loss function. In this paper, instead of expressing accurate oriented information by the rotation angle, we propose a novel projection-based method that employs six-parameter to describe rotated bounding box: two points position on the projected line(one center point and one chosen vertex), a projection ratio and a quadrant label, named as ProjBB. ProjBB is angle-free and has three main advantages: i) More accurate angle distance without periodicity; ii) Projection ratio which is aspect ratio sensitive; iii) Theoretical uniqueness guarantee. Extensive experiments on two large-scale public aerial image datasets(i.e. DOTA and HRSC2016) and the scene text dataset ICDAR2015 show the competitive results of our proposed approach.

show abstract

Section: Regression Problems In Angle-based Methodsmentioning

confidence: 99%

Section: Rotation Detector Built On Retinanetmentioning

confidence: 99%

Section: Rotation Detector Built On Retinanetmentioning

confidence: 99%

Section: E Comparison and Summarymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Bounding Box Projection for Regression Uncertainty in Oriented Object Detection

Xiang

Tang³

et al. 2021

IEEE Access

View full text Add to dashboard Cite

show abstract

Jensen–Shannon Divergence You Only Look Once: A Real‐Time Robotic Grasp Detection Network

Han,

2024

Advanced Intelligent Systems

View full text Add to dashboard Cite

In this article, the arbitrary‐oriented object detection problem with application in robotic grasping is addressed. A novel Jensen–Shannon divergence (JSD)– You Only Look Once (YOLO) model is proposed, which enables real‐time grasp detection with high performance. The one‐stage object detection network YOLOv5 is modified with a decoupled head, which solves the angle classification problem and rectangle parameter regression problem separately, such that the YOLOv5 network is applicable for robotic grasping and the detection accuracy is significantly improved. A circular smooth label angle classification method is proposed to tackle the boundary discontinuity problem in angle regression, and the periodicity of the angle prediction is guaranteed. A novel Jensen–Shannon intersection of union is designed to calculate the intersection over union of oriented rectangles, which aims to better measure the discrepancies between the prediction and the ground truth and to avoid the singularity problem when two rectangles are not overlapped. Extensive evaluation on the Cornell and visual manipulation relationship dataset datasets demonstrates the effectiveness of the JSD–YOLO model in general robotic grasp operations, with 99.7% and 95.7% image‐wise split accuracy, respectively.

show abstract

Toward automatic beam angle selection for pencil‐beam scanning proton liver treatments: A deep learning–based approach

Kaderka

Liu²,

Liu³

et al. 2022

Medical Physics

View full text Add to dashboard Cite

Background Dose deposition characteristics of proton radiation can be advantageous over photons. Proton treatment planning, however, poses additional challenges for the planners. Proton therapy is usually delivered with only a small number of beam angles, and the quality of a proton treatment plan is largely determined by the beam angles employed. Finding the optimal beam angles for a proton treatment plan requires time and experience, motivating the investigation of automatic beam angle selection methods. Purpose A deep learning–based approach to automatic beam angle selection is proposed for the proton pencil‐beam scanning treatment planning of liver lesions. Methods We cast beam‐angle selection as a multi‐label classification problem. To account for angular boundary discontinuity, the underlying convolution neural network is trained with the proposed Circular Earth Mover's Distance–based regularization and multi‐label circular–smooth label technique. Furthermore, an analytical algorithm emulating proton treatment planners’ clinical practice is employed in post‐processing to improve the output of the model. Forty‐nine patients that received proton liver treatments between 2017 and 2020 were randomly divided into training (n = 31), validation (n = 7), and test sets (n = 11). AI‐selected beam angles were compared with those angles selected by human planners, and the dosimetric outcome was investigated by creating plans using knowledge‐based treatment planning. Results For 7 of the 11 cases in the test set, AI‐selected beam angles agreed with those chosen by human planners to within 20° (median angle difference = 10°; mean = 18.6°). Moreover, out of the total 22 beam angles predicted by the model, 15 (68%) were within 10° of the human‐selected angles. The high correlation in beam angles resulted in comparable dosimetric statistics between proton treatment plans generated using AI‐ and human‐selected angles. For the cases with beam angle differences exceeding 20°, the dosimetric analysis showed similar plan quality although with different emphases on organ‐at‐risk sparing. Conclusions This pilot study demonstrated the feasibility of a novel deep learning–based beam angle selection technique. Testing on liver cancer patients showed that the resulting plans were clinically viable with comparable dosimetric quality to those using human‐selected beam angles. In tandem with auto‐contouring and knowledge‐based treatment planning tools, the proposed model could represent a pathway for nearly fully automated treatment planning in proton therapy.

show abstract

Arbitrary-Oriented Object Detection with Circular Smooth Label

Cited by 429 publications

References 50 publications

Bounding Box Projection for Regression Uncertainty in Oriented Object Detection

Bounding Box Projection for Regression Uncertainty in Oriented Object Detection

Jensen–Shannon Divergence You Only Look Once: A Real‐Time Robotic Grasp Detection Network

Toward automatic beam angle selection for pencil‐beam scanning proton liver treatments: A deep learning–based approach

Contact Info

Product

Resources

About