G-Rep: Gaussian Representation for Arbitrary-Oriented Object Detection. Hou, L., Lu, K., Yang, X., Li, Y., & Xue, J. Remote Sensing, 15(3):1-21, 2023. ![G-Rep: Gaussian Representation for Arbitrary-Oriented Object Detection [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Paper doi abstract bibtex Typical representations for arbitrary-oriented object detection tasks include the oriented bounding box (OBB), the quadrilateral bounding box (QBB), and the point set (PointSet). Each representation encounters problems that correspond to its characteristics, such as boundary discontinuity, square-like problems, representation ambiguity, and isolated points, which lead to inaccurate detection. Although many effective strategies have been proposed for various representations, there is still no unified solution. Current detection methods based on Gaussian modeling have demonstrated the possibility of resolving this dilemma; however, they remain limited to OBB. To go further, in this paper, we propose a unified Gaussian representation called G-Rep to construct Gaussian distributions for OBB, QBB, and PointSet, which achieves a unified solution to various representations and problems. Specifically, PointSet- or QBB-based object representations are converted into Gaussian distributions and their parameters are optimized using the maximum likelihood estimation algorithm. Then, three optional Gaussian metrics are explored to optimize the regression loss of the detector because of their excellent parameter optimization mechanisms. Furthermore, we also use Gaussian metrics for sampling to align label assignment and regression loss. Experimental results obtained on several publicly available datasets, such as DOTA, HRSC2016, UCAS-AOD, and ICDAR2015, show the excellent performance of the proposed method for arbitrary-oriented object detection.
@article{
title = {G-Rep: Gaussian Representation for Arbitrary-Oriented Object Detection},
type = {article},
year = {2023},
keywords = {Gaussian metrics,Gaussian representation,arbitrary-oriented object detection,convolutional neural networks},
pages = {1-21},
volume = {15},
id = {545989fb-8ae1-39c1-aadd-31d1562b6398},
created = {2024-03-27T10:37:21.871Z},
file_attached = {true},
profile_id = {bfbbf840-4c42-3914-a463-19024f50b30c},
group_id = {5ec9cc91-a5d6-3de5-82f3-3ef3d98a89c1},
last_modified = {2024-04-08T06:04:34.526Z},
read = {true},
starred = {false},
authored = {false},
confirmed = {true},
hidden = {false},
folder_uuids = {df28411a-ed7f-4991-8358-d39685eb4bf0},
private_publication = {false},
abstract = {Typical representations for arbitrary-oriented object detection tasks include the oriented bounding box (OBB), the quadrilateral bounding box (QBB), and the point set (PointSet). Each representation encounters problems that correspond to its characteristics, such as boundary discontinuity, square-like problems, representation ambiguity, and isolated points, which lead to inaccurate detection. Although many effective strategies have been proposed for various representations, there is still no unified solution. Current detection methods based on Gaussian modeling have demonstrated the possibility of resolving this dilemma; however, they remain limited to OBB. To go further, in this paper, we propose a unified Gaussian representation called G-Rep to construct Gaussian distributions for OBB, QBB, and PointSet, which achieves a unified solution to various representations and problems. Specifically, PointSet- or QBB-based object representations are converted into Gaussian distributions and their parameters are optimized using the maximum likelihood estimation algorithm. Then, three optional Gaussian metrics are explored to optimize the regression loss of the detector because of their excellent parameter optimization mechanisms. Furthermore, we also use Gaussian metrics for sampling to align label assignment and regression loss. Experimental results obtained on several publicly available datasets, such as DOTA, HRSC2016, UCAS-AOD, and ICDAR2015, show the excellent performance of the proposed method for arbitrary-oriented object detection.},
bibtype = {article},
author = {Hou, Liping and Lu, Ke and Yang, Xue and Li, Yuqiu and Xue, Jian},
doi = {10.3390/rs15030757},
journal = {Remote Sensing},
number = {3}
}
Downloads: 0
{"_id":"H7Fm4R69NAJLFWtbr","bibbaseid":"hou-lu-yang-li-xue-grepgaussianrepresentationforarbitraryorientedobjectdetection-2023","author_short":["Hou, L.","Lu, K.","Yang, X.","Li, Y.","Xue, J."],"bibdata":{"title":"G-Rep: Gaussian Representation for Arbitrary-Oriented Object Detection","type":"article","year":"2023","keywords":"Gaussian metrics,Gaussian representation,arbitrary-oriented object detection,convolutional neural networks","pages":"1-21","volume":"15","id":"545989fb-8ae1-39c1-aadd-31d1562b6398","created":"2024-03-27T10:37:21.871Z","file_attached":"true","profile_id":"bfbbf840-4c42-3914-a463-19024f50b30c","group_id":"5ec9cc91-a5d6-3de5-82f3-3ef3d98a89c1","last_modified":"2024-04-08T06:04:34.526Z","read":"true","starred":false,"authored":false,"confirmed":"true","hidden":false,"folder_uuids":"df28411a-ed7f-4991-8358-d39685eb4bf0","private_publication":false,"abstract":"Typical representations for arbitrary-oriented object detection tasks include the oriented bounding box (OBB), the quadrilateral bounding box (QBB), and the point set (PointSet). Each representation encounters problems that correspond to its characteristics, such as boundary discontinuity, square-like problems, representation ambiguity, and isolated points, which lead to inaccurate detection. Although many effective strategies have been proposed for various representations, there is still no unified solution. Current detection methods based on Gaussian modeling have demonstrated the possibility of resolving this dilemma; however, they remain limited to OBB. To go further, in this paper, we propose a unified Gaussian representation called G-Rep to construct Gaussian distributions for OBB, QBB, and PointSet, which achieves a unified solution to various representations and problems. Specifically, PointSet- or QBB-based object representations are converted into Gaussian distributions and their parameters are optimized using the maximum likelihood estimation algorithm. Then, three optional Gaussian metrics are explored to optimize the regression loss of the detector because of their excellent parameter optimization mechanisms. Furthermore, we also use Gaussian metrics for sampling to align label assignment and regression loss. Experimental results obtained on several publicly available datasets, such as DOTA, HRSC2016, UCAS-AOD, and ICDAR2015, show the excellent performance of the proposed method for arbitrary-oriented object detection.","bibtype":"article","author":"Hou, Liping and Lu, Ke and Yang, Xue and Li, Yuqiu and Xue, Jian","doi":"10.3390/rs15030757","journal":"Remote Sensing","number":"3","bibtex":"@article{\n title = {G-Rep: Gaussian Representation for Arbitrary-Oriented Object Detection},\n type = {article},\n year = {2023},\n keywords = {Gaussian metrics,Gaussian representation,arbitrary-oriented object detection,convolutional neural networks},\n pages = {1-21},\n volume = {15},\n id = {545989fb-8ae1-39c1-aadd-31d1562b6398},\n created = {2024-03-27T10:37:21.871Z},\n file_attached = {true},\n profile_id = {bfbbf840-4c42-3914-a463-19024f50b30c},\n group_id = {5ec9cc91-a5d6-3de5-82f3-3ef3d98a89c1},\n last_modified = {2024-04-08T06:04:34.526Z},\n read = {true},\n starred = {false},\n authored = {false},\n confirmed = {true},\n hidden = {false},\n folder_uuids = {df28411a-ed7f-4991-8358-d39685eb4bf0},\n private_publication = {false},\n abstract = {Typical representations for arbitrary-oriented object detection tasks include the oriented bounding box (OBB), the quadrilateral bounding box (QBB), and the point set (PointSet). Each representation encounters problems that correspond to its characteristics, such as boundary discontinuity, square-like problems, representation ambiguity, and isolated points, which lead to inaccurate detection. Although many effective strategies have been proposed for various representations, there is still no unified solution. Current detection methods based on Gaussian modeling have demonstrated the possibility of resolving this dilemma; however, they remain limited to OBB. To go further, in this paper, we propose a unified Gaussian representation called G-Rep to construct Gaussian distributions for OBB, QBB, and PointSet, which achieves a unified solution to various representations and problems. Specifically, PointSet- or QBB-based object representations are converted into Gaussian distributions and their parameters are optimized using the maximum likelihood estimation algorithm. Then, three optional Gaussian metrics are explored to optimize the regression loss of the detector because of their excellent parameter optimization mechanisms. Furthermore, we also use Gaussian metrics for sampling to align label assignment and regression loss. Experimental results obtained on several publicly available datasets, such as DOTA, HRSC2016, UCAS-AOD, and ICDAR2015, show the excellent performance of the proposed method for arbitrary-oriented object detection.},\n bibtype = {article},\n author = {Hou, Liping and Lu, Ke and Yang, Xue and Li, Yuqiu and Xue, Jian},\n doi = {10.3390/rs15030757},\n journal = {Remote Sensing},\n number = {3}\n}","author_short":["Hou, L.","Lu, K.","Yang, X.","Li, Y.","Xue, J."],"urls":{"Paper":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c/file/5a4b30fc-cbe8-cc7f-79d2-e6f7f7d80096/remotesensing_15_00757.pdf.pdf"},"biburl":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c","bibbaseid":"hou-lu-yang-li-xue-grepgaussianrepresentationforarbitraryorientedobjectdetection-2023","role":"author","keyword":["Gaussian metrics","Gaussian representation","arbitrary-oriented object detection","convolutional neural networks"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"article","biburl":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c","dataSources":["2252seNhipfTmjEBQ"],"keywords":["gaussian metrics","gaussian representation","arbitrary-oriented object detection","convolutional neural networks"],"search_terms":["rep","gaussian","representation","arbitrary","oriented","object","detection","hou","lu","yang","li","xue"],"title":"G-Rep: Gaussian Representation for Arbitrary-Oriented Object Detection","year":2023}