• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Computer Animation and Virtual Worlds >ASFNet: Adaptive multiscale segmentation fusion network for real-time semantic segmentation
【24h】

ASFNet: Adaptive multiscale segmentation fusion network for real-time semantic segmentation

机译：ASFNET：自适应多尺度分割融合网络，用于实时语义分割

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Recently, the development of deep learning has facilitated continuous progress in the field of computer vision. Pixel-level semantic segmentation serves as a fundamental task in computer vision. It achieves significant results by connecting wider and deeper backbone networks and building fine-grained segmentation heads. However, applications such as self-driving cars are more critical to the computational speed of the algorithms. The trade-off between accuracy and real-time performance of existing algorithms is still a challenging task. To address this challenge, this article proposes an adaptive multiscale segmentation fusion network to fuse multiscale contextual, which designs an adaptive multiscale segmentation fusion module based on an attention mechanism. Using segmentation fusion instead of feature fusion, the multiscale segmentation results are aggregated to obtain more precise segmentation results. The final results achieved 70.9% mIoU of accuracy in the Cityspace test set, processing images at 61 FPS when the input is 1024 x 2048. In addition, when adjusting the input size to 512 x 1024, the images are processed at 185 FPS.
机译：最近，深度学习的发展促进了计算机愿景领域的持续进步。像素级语义分割用作计算机视觉中的基本任务。它通过连接更广泛和更深的骨干网络和建筑细粒细分头来实现显着的结果。然而，自动驾驶汽车等应用对算法的计算速度更为重要。现有算法的准确性和实时性能之间的权衡仍然是一个具有挑战性的任务。为了解决这一挑战，本文提出了一种自适应多尺度分段融合网络，用于熔断多尺度上下文，这为基于注意机制设计自适应多尺度分段融合模块。使用分段融合而不是特征融合，聚合多尺度分段结果以获得更精确的分段结果。最终结果在CitySpace测试集中实现了70.9％的精度，当输入是1024 x 2048时，在61 fps下处理图像。另外，在将输入大小调整为512 x 1024时，图像将以185 fps处理。

著录项

  • 来源
    《Computer Animation and Virtual Worlds》 |2021年第4期|e2022.1-e2022.11|共11页
  • 作者

    Zha Hengfeng; Liu Rui; Yang Xin; Zhou Dongsheng; Zhang Qiang; Wei Xiaopeng;

  • 作者单位

    Dalian Univ Sch Software Engn Natl & Local Joint Engn Lab Comp Aided Design Dalian 116622 Peoples R China;

    Dalian Univ Sch Software Engn Natl & Local Joint Engn Lab Comp Aided Design Dalian 116622 Peoples R China;

    Dalian Univ Technol Sch Comp Sci & Technol Dalian Peoples R China;

    Dalian Univ Sch Software Engn Natl & Local Joint Engn Lab Comp Aided Design Dalian 116622 Peoples R China|Dalian Univ Technol Sch Comp Sci & Technol Dalian Peoples R China;

    Dalian Univ Sch Software Engn Natl & Local Joint Engn Lab Comp Aided Design Dalian 116622 Peoples R China|Dalian Univ Technol Sch Comp Sci & Technol Dalian Peoples R China;

    Dalian Univ Technol Sch Comp Sci & Technol Dalian Peoples R China;

  • 收录信息
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

    computer vision; multiscale fusion; real-time semantic segmentation; segmentation fusion;

    机译：计算机愿景;多尺度融合;实时语义分割;分割融合;

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号