Oriented objects as pairs of middle lines

Wei, Haoran; Zhang, Yue; Chang, Zhonghan; Li, Hao; Wang, Hongqi; Sun, Xian

doi:10.1016/j.isprsjprs.2020.09.022

Cited by 197 publications

(63 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, the values of these two parameters α = 0.25 and γ = 2 were set to zeropoint-two-five and two empirically. Meanwhile, as shown in Table 1, we set the value of λ = {0.01, 0.1, 0.2, 0.5, 0.75, 1} in (7) and achieved the highest mAP of 96.33% when λ = 0.5. Therefore, we chose 0.5 as the λ value for the best performance.…”

Section: Experimental Details and Network Inference 421 Experimental Detailsmentioning

confidence: 99%

“…Yi et al [6] represented the objects in remote sensing via the center keypoints and regressed the box boundary-aware vectors (BBAVectors) to locate the AOBB targets. O 2 -DNet [7] detected the oriented targets in remote sensing images by predicting a pair of middle lines inside each bounding box. To the best of our knowledge, the above anchor-free oriented object detectors can be simplified into two typical models: (1) directly or indirectly regress the coordinates of the four vertices {V i = (x i , y i )|i = 1, 2, 3, 4} of AOBB; (2) directly or indirectly regress the center coordinate (x c , y c ), the scale of the AOBB, such as the lengths of the long and short sides (w, h), and the orientation θ of the target.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Multi-Sector Oriented Object Detector for Accurate Localization in Optical Remote Sensing Images

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

Oriented object detection in optical remote sensing images (ORSIs) is a challenging task since the targets in ORSIs are displayed in an arbitrarily oriented manner and on small scales, and are densely packed. Current state-of-the-art oriented object detection models used in ORSIs primarily evolved from anchor-based and direct regression-based detection paradigms. Nevertheless, they still encounter a design difficulty from handcrafted anchor definitions and learning complexities in direct localization regression. To tackle these issues, in this paper, we proposed a novel multi-sector oriented object detection framework called MSO2-Det, which quantizes the scales and orientation prediction of targets in ORSIs via an anchor-free classification-to-regression approach. Specifically, we first represented the arbitrarily oriented bounding box as four scale offsets and angles in four quadrant sectors of the corresponding Cartesian coordinate system. Then, we divided the scales and angle space into multiple discrete sectors and obtained more accurate localization information by a coarse-granularity classification to fine-grained regression strategy. In addition, to decrease the angular-sector classification loss and accelerate the network’s convergence, we designed a smooth angular-sector label (SASL) that smoothly distributes label values with a definite tolerance radius. Finally, we proposed a localization-aided detection score (LADS) to better represent the confidence of a detected box by combining the category-classification score and the sector-selection score. The proposed MSO2-Det achieves state-of-the-art results on three widely used benchmarks, including the DOTA, HRSC2016, and UCAS-AOD data sets.

show abstract

Section: Experimental Details and Network Inference 421 Experimental Detailsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Multi-Sector Oriented Object Detector for Accurate Localization in Optical Remote Sensing Images

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Object detection is an important task in the remote sensing image interpretation. With the application of deep learning in computer vision [1][2][3][4], an increasing number of object detection methods based on convolutional neural networks (CNNs) [5][6][7][8][9][10][11] have been proposed to achieve good performance. However, fully supervised object detection methods require a large number of samples with instance-level labels.…”

Section: Introductionmentioning

confidence: 99%

Dynamic Pseudo-Label Generation for Weakly Supervised Object Detection in Remote Sensing Images

Wang

Qian

et al. 2021

Remote Sensing

Self Cite

View full text Add to dashboard Cite

In recent years, fully supervised object detection methods in remote sensing images with good performance have been developed. However, this approach requires a large number of instance-level annotated samples that are relatively expensive to acquire. Therefore, weakly supervised learning using only image-level annotations has attracted much attention. Most of the weakly supervised object detection methods are based on multi-instance learning methods, and their performance depends on the process of scoring the candidate region proposals during training. In this process, the use of only image-level labels for supervision usually cannot obtain optimal results due to the lack of location information of the object. To address the above problem, a dynamic sample pseudo-label generation framework is proposed to generate pseudo-labels for each proposal without additional annotations. First, we propose the pseudo-label generation algorithm (PLG) to generate the category labels of the proposal by using the localization information of the object. Specifically, we propose to use the pixel average of the object’s localization map in the proposal as the proposal category confidence and calculate the pseudo-label by comparing the proposal category confidence with the preset threshold. In addition, an effective adaptive threshold selection strategy is designed to eliminate the effect of different category shape differences in computing sample pseudo-labels. Comparative experiments on the NWPU VHR-10 dataset demonstrate that our method can significantly improve the detection performance compared to existing methods.

show abstract

“…Because the output prediction map is also sparse and few non-empty grids are assigned a foreground label makes the anchor-based detector training difficult in the training step, we predict objects by grid-wise foreground segmentation rather than the anchor-based method. Inspired by [10], we represent objects as two mutually perpendicular lines across the foreground grids. We regress endpoints of two lines, and these values are used to calculate the center, dimension, and orientation of objects indirectly.…”

Section: Introductionmentioning

confidence: 99%

“…We predict the score of every direction and get the bounding box with the maximum direction score. In [10], besides the endpoints regression loss, the collinear loss, and the vertical loss are introduced to constrain the endpoints of every line to be collinear and the two lines to be perpendicular to each other. To regress the bounding box better, we design a serial of loss functions for regressing objects' center, dimension, and orientation.…”

Section: Introductionmentioning

confidence: 99%

Real-Time 3D Object Detection From Point Cloud Through Foreground Segmentation

et al. 2021

View full text Add to dashboard Cite

This paper aims to apply real-time light-weight high-precision 3D detection for autonomous driving. We propose LIDAR-based 3D object detection based on foreground segmentation using a fully sparse convolutional network (FS 2 3D). We design a sparse convolutional backbone network and a sparse convolutional detection head to efficiently use the computing and memory resources and accelerate the inference. Instead of using the anchor-based method, we convert the detection problem into a foreground segmentation problem on a bird's-eye view. The sparse convolutional detection head predicts the objectness and bounding box on each active point on the sparse feature map. We design a new oriented bounding box coding method and corresponding loss functions. We predict the endpoints of two mutually perpendicular lines that pass through the foreground active points and indirectly predict the objects' oriented bounding box from these four endpoints. We use the indirectly calculated object center, size, and orientation as inputs of loss functions in the training step. Experiments on the KITTI dataset show that the sparse backbone network we designed is 2.2 times faster and 18.4 times fewer FLOPs than the dense backbone network. The average improvement of the loss functions based on the bounding box code is 1.1% and 0.8% on the BEV and 3D detection, respectively, compared to no addition of these losses. Moreover, FS 2 3D outperforms the state-of-the-art LIDAR-based method in speed and precision for both cars and cyclists.INDEX TERMS 3D object Detection, LIDAR, real-time, sparse convolutional neural network.

show abstract

Oriented objects as pairs of middle lines

Cited by 197 publications

References 38 publications

Multi-Sector Oriented Object Detector for Accurate Localization in Optical Remote Sensing Images

Multi-Sector Oriented Object Detector for Accurate Localization in Optical Remote Sensing Images

Dynamic Pseudo-Label Generation for Weakly Supervised Object Detection in Remote Sensing Images

Real-Time 3D Object Detection From Point Cloud Through Foreground Segmentation

Contact Info

Product

Resources

About