Bounding Box Regression With Uncertainty for Accurate Object Detection

He, Yihui; Zhu, Chenchen; Wang, Jianren; Savvides, Marios; Zhang, Xiangyu

doi:10.1109/cvpr.2019.00300

Cited by 502 publications

(320 citation statements)

References 47 publications

Supporting

Mentioning

318

Contrasting

Order By: Relevance

“…The class which has more samples in a dataset or mini batch during training in the context of class imbalance. [17] and in KL loss [54] for Smooth L1 Loss), while some methods such as GIoU Loss [55] directly predict the bounding box coordinates. For the sake of clarity, we usex to denote the regression loss input for any method.…”

Section: Over-represented Classmentioning

confidence: 99%

See 1 more Smart Citation

Imbalance Problems in Object Detection: A Review

Öksüz

Cam

Kalkan

et al. 2021

IEEE Trans. Pattern Anal. Mach. Intell.

412

171

View full text Add to dashboard Cite

In this paper, we present a comprehensive review of the imbalance problems in object detection. To analyze the problems in a systematic manner, we introduce two taxonomies; one for the problems and the other for the proposed solutions. Following the taxonomy for the problems, we discuss each problem in depth and present a unifying yet critical perspective on the solutions in the literature. In addition, we identify major open issues regarding the existing imbalance problems as well as imbalance problems that have not been discussed before. Moreover, in order to keep our review up to date, we provide an accompanying webpage which categorizes papers addressing imbalance problems, according to our problem-based taxonomy. Researchers can track newer studies on this webpage available at: https://github.com/kemaloksuz/ObjectDetectionImbalance.

show abstract

Section: Over-represented Classmentioning

confidence: 99%

“…Recent regression loss functions have different motivations: (i) Balanced L1 Loss [29] increases the contribution of the inliers. (ii) KL Loss [54] is motivated from the ambiguity of the positive samples. (iii) GIoU Loss [55] has the motive to use a performance metric as a loss function.…”

Section: A Regression Loss With Many Aspectsmentioning

confidence: 99%

Imbalance Problems in Object Detection: A Review

Öksüz

Cam

Kalkan

et al. 2021

IEEE Trans. Pattern Anal. Mach. Intell.

412

171

View full text Add to dashboard Cite

show abstract

“…Zhu et al [41] enhanced the anchor design for small objects. He et al [14] modeled the bounding The idea of anchor-free detection is not new. Dense-Box [15] first proposed a unified end-to-end fully convolutional framework that directly predicted bounding boxes.…”

Section: Related Workmentioning

confidence: 99%

Feature Selective Anchor-Free Module for Single-Shot Object Detection

Zhu

Savvides

2019

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

Self Cite

857

372

View full text Add to dashboard Cite

person 0.90 person 0.83 person 0.90 person 0.69 person 0.68 person 0.57 person 0.69 skis 0.59 person 0.53 person 0.53 a: RetinaNet (anchor-based, ResNeXt-101) b: Ours (anchor-based + FSAF, ResNet-50) Figure 1: Qualitative results of the anchor-based RetinaNet [22] using powerful ResNeXt-101 (left) and our detector with additional FSAF module using just ResNet-50 (right) under the same training and testing scale. Our FSAF module helps detecting hard objects like tiny person and flat skis with a less powerful backbone network. See Figure 7 for more examples. AbstractWe motivate and present feature selective anchor-free (FSAF) module, a simple and effective building block for single-shot object detectors. It can be plugged into singleshot detectors with feature pyramid structure. The FSAF module addresses two limitations brought up by the conventional anchor-based detection: 1) heuristic-guided feature selection; 2) overlap-based anchor sampling. The general concept of the FSAF module is online feature selection applied to the training of multi-level anchor-free branches. Specifically, an anchor-free branch is attached to each level of the feature pyramid, allowing box encoding and decoding in the anchor-free manner at an arbitrary level. During training, we dynamically assign each instance to the most suitable feature level. At the time of inference, the FSAF module can work jointly with anchor-based branches by outputting predictions in parallel. We instantiate this concept with simple implementations of anchor-free branches and online feature selection strategy. Experimental re-sults on the COCO detection track show that our FSAF module performs better than anchor-based counterparts while being faster. When working jointly with anchor-based branches, the FSAF module robustly improves the baseline RetinaNet by a large margin under various settings, while introducing nearly free inference overhead. And the resulting best model can achieve a state-of-the-art 44.6% mAP, outperforming all existing single-shot detectors on COCO.

show abstract

“…To the best of our knowledge, the most similar work to our approach is the recent Softer-NMS [7] that also uses KL divergence for bounding box localization. However, the motivation of Softer-NMS is to model the location uncertainty of each corner of proposed bounding boxes through 1D Gaussian distribution.…”

Section: Bounding Ellipse and Kl Divergencementioning

confidence: 99%

Detecting Lesion Bounding Ellipses with Gaussian Proposal Networks

Li¹

2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Lesions characterized by computed tomography (CT) scans, are arguably often elliptical objects. However, current lesion detection systems are predominantly adopted from the popular Region Proposal Networks (RPNs) [21] that only propose bounding boxes without fully leveraging the elliptical geometry of lesions. In this paper, we present Gaussian Proposal Networks (GPNs), a novel extension to RPNs, to detect lesion bounding ellipses. Instead of directly regressing the rotation angle of the ellipse as the common practice, GPN represents bounding ellipses as 2D Gaussian distributions on the image plain and minimizes the Kullback-Leibler (KL) divergence between the proposed Gaussian and the ground truth Gaussian for object localization. We show the KL divergence loss approximately incarnates the regression loss in the RPN framework when the rotation angle is 0. Experiments on the DeepLesion [27] dataset show that GPN significantly outperforms RPN for lesion bounding ellipse detection thanks to lower localization error. GPN is open sourced at https://github.com/baidu-research/GPN.

show abstract

Bounding Box Regression With Uncertainty for Accurate Object Detection

Cited by 502 publications

References 47 publications

Imbalance Problems in Object Detection: A Review

Imbalance Problems in Object Detection: A Review

Feature Selective Anchor-Free Module for Single-Shot Object Detection

Detecting Lesion Bounding Ellipses with Gaussian Proposal Networks

Contact Info

Product

Resources

About