Z
Zehua Huang
Publications - 18
Citations - 2094
Zehua Huang is an academic researcher. The author has contributed to research in topics: Segmentation & Upsampling. The author has an hindex of 8, co-authored 17 publications receiving 1491 citations.
Papers
More filters
Proceedings ArticleDOI
Understanding Convolution for Semantic Segmentation
TL;DR: DUC is designed to generate pixel-level prediction, which is able to capture and decode more detailed information that is generally missing in bilinear upsampling, and a hybrid dilated convolution (HDC) framework in the encoding phase is proposed.
Posted Content
Understanding Convolution for Semantic Segmentation
TL;DR: Zhang et al. as mentioned in this paper design dense upsampling convolution (DUC) to generate pixel-level prediction, which is able to capture and decode more detailed information that is generally missing in bilinear upampling.
Patent
System and method for semantic segmentation using dense upsampling convolution (DUC)
TL;DR: In this article, a system and method for semantic segmentation using dense upsampling convolution (DUC) is described, in which the label map is divided into equal subparts, which have the same height and width as the feature map.
Patent
System and method for semantic segmentation using hybrid dilated convolution (HDC)
TL;DR: In this article, a method for semantic segmentation using hybrid dilated convolution (HDC) is described. But the method is not suitable for image segmentation with multiple convolutional layers.
Patent
System and method for occluding contour detection
TL;DR: In this article, a system method for occluding contour detection using a fully convolutional neural network (FCN) is described. But the method is limited to the detection of objects and object instances detected in the input image.