DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Yoo, Chang Dong | - |
dc.contributor.advisor | 유창동 | - |
dc.contributor.author | Jang, Hyungjun | - |
dc.date.accessioned | 2021-05-13T19:34:09Z | - |
dc.date.available | 2021-05-13T19:34:09Z | - |
dc.date.issued | 2020 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=911398&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/284768 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 전기및전자공학부, 2020.2,[v, 28 p. :] | - |
dc.description.abstract | This paper considers an architecture, referred to as RankNet, for improving object detection by addressing limitations in existing object detection methods as follows: (1) reweights the important samples considering the IoU and classification score of each positive and negative bounding box, (2) define a ranking score to measure the similarity between ground truth and predicted bounding box to indicate the ranks of bounding boxes for NMS procedure, and a metric called EIoU was proposed to calculate the ranking score. Extensive experiments on the standard COCO dataset show the effectiveness of the proposed method in multiple evaluation metrics, including multiple threshold AP metric. In particular, the proposed RankNet improves the AP of the bounding box 1.2, 2.3, 0.9 points compared to the Fast/Faster and Cascade R-CNN baseline, respectively. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Object Detection▼aBox Ranking▼aNMS▼aDeep Learning | - |
dc.subject | 객체 검출▼a상자 순위▼a상자 억제▼a심층 학습 | - |
dc.title | RankNet for accurate object detection with ranking bounding boxes | - |
dc.title.alternative | 박스 회귀 순위를 통해 정확한 객체 검출을 위한 순위 네트워크 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :전기및전자공학부, | - |
dc.contributor.alternativeauthor | 장현준 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.