1. 首页
  2. 人工智能
  3. 深度学习
  4. Tencent CNN in MRF: Video Object Segmentation SpatioTemporal MRF

Tencent CNN in MRF: Video Object Segmentation SpatioTemporal MRF

上传者: 2019-09-14 16:34:50上传 PDF文件 2.51MB 热度 26次
Thispaperaddressestheproblemofvideoobjectsegmentation,wheretheinitialobjectmaskisgiveninthefirstframeofaninputvideo.WeproposeanovelspatiotemporalMarkovRandomField(MRF)modeldefinedoverpixelstohandlethisproblem.UnlikeconventionalMRFmodels,thespatialdependenciesamongpixelsinourmodela
用户评论