G
Geonseok Seo
Researcher at Seoul National University
Publications - 7
Citations - 34
Geonseok Seo is an academic researcher from Seoul National University. The author has contributed to research in topics: Object detection & Minimum bounding box. The author has an hindex of 3, co-authored 7 publications receiving 22 citations.
Papers
More filters
Posted Content
C3: Concentrated-Comprehensive Convolution and its application to semantic segmentation
TL;DR: A new block called Concentrated-Comprehensive Convolution (C3) is proposed which applies the asymmetric convolutions before the depth-wise separable dilated Convolution to compensate for the information loss due to dilated convolution.
Posted Content
Concentrated-Comprehensive Convolutions for lightweight semantic segmentation
TL;DR: A new block of Concentrated-Comprehensive Convolution (CCC) is proposed which takes both advantages of the dilated Convolution and the depth-wise separable convolution and can successfully replace dilated convolutions on ImageNet classification task.
Proceedings ArticleDOI
Kl-Divergence-Based Region Proposal Network For Object Detection
TL;DR: Zhang et al. as discussed by the authors proposed a new region proposal learning method that considers the bounding box offset's uncertainty in the objectness score, which can improve the performance of the existing region proposal network.
Journal Article
Density-Based Object Detection: Learning Bounding Boxes without Ground Truth Assignment
TL;DR: This paper reformulate the multi-object detection task as a problem of density estimation of bounding boxes as well as proposing a novel network for object detection called Mixture Density Object Detector (MDOD), and the corresponding objective function for the density-estimation-based training.
Posted Content
Training Multi-Object Detector by Estimating Bounding Box Distribution for Input Image
TL;DR: Yoo et al. as mentioned in this paper reformulated the multi-object detection task as a problem of density estimation of bounding boxes and proposed a mixture density object detector (MDOD) to solve the problem.