本文主要有3个贡献：1、提出了PSPNet，在FCN中嵌入了不同场景的上下文特征；2、我们基于深度监督的损失为deep ResNet开发了有效的优化策略；3、我们构建了一个用于最新场景解析和语义分割的实用系统，其中包括了所有关键的实现细节。 语义分割任务在当时的来源是把图像分类的模型，全连接层改成全卷积层。 扩大感受野常用的方法是空洞卷积，本文的baseline network就 …

2016年12月4日 · In this paper, we exploit the capability of global context information by different-region-based context aggregation through our pyramid pooling module together with the proposed pyramid scene parsing network (PSPNet).

2024年2月5日 · PSPNet, an acronym for Pyramid Scene Parsing Network, constitutes a profound Deep Learning model meticulously crafted for pixel-wise semantic segmentation of images. Developed by Heng Shuang Zhao et al. in 2017, PSPNet adeptly tackles the challenges associated with capturing contextual information across diverse scales within an image.

2017年11月9日 · In this paper, we exploit the capability of global context information by different-region-based context aggregation through our pyramid pooling module together with the proposed pyramid scene parsing network (PSPNet).

PSPNet, or Pyramid Scene Parsing Network, is a semantic segmentation model that utilises a pyramid parsing module that exploits global context information by different-region based context aggregation.

In this paper, we exploit the capability of global context information by different-region-based context aggregation through our pyramid pooling module together with the proposed pyramid scene parsing network (PSPNet).

2020年5月15日 · This repository is a PyTorch implementation for semantic segmentation / scene parsing. The code is easy to use for training and testing on various datasets. The codebase mainly uses ResNet50/101/152 as backbone and can be easily adapted to other basic classification structures.

Highly optimized PyTorch codebases available for semantic segmentation in repo: semseg, including full training and testing codes for PSPNet and PSANet.

In this guide, we will mainly focus on Pyramid scene parsing network (PSPNet) [1] which is one of the most well-recognized image segmentation algorithms as it won ImageNet Scene Parsing Challenge 2016 and its paper is highly cited by the computer vision community.

Using pretrained weights for extractors - improved quality and convergence dramatically.