search

oravus / lostX

(RSS 2018) LoST - Visual Place Recognition using Visual Semantics for Opposite Viewpoints across Day and Night
MIT License
76 stars 11 forks source link

pretrained model on SYNTHIA dataset #2

Closed BYQ-ZJU closed 4 years ago

BYQ-ZJU commented 4 years ago

would you please upload trained model on SYNTHIA?

oravus commented 4 years ago

Hi, we used RefineNet trained on the Cityscapes dataset for semantic segmentation. The trained models for that can be downloaded by visiting https://github.com/guosheng/refinenet

BYQ-ZJU commented 4 years ago

Hi, we used RefineNet trained on the Cityscapes dataset for semantic segmentation. The trained models for that can be downloaded by visiting https://github.com/guosheng/refinenet thank you for the reply.