RichExplor / CNN_VINS

using SuperPoint as visual front to VINS-Mono
91 stars 9 forks source link

我使用eric-yyjau / pytorch-superpoint的方法训练模型,发现效果很差 #4

Open FeiXie8 opened 2 years ago

FeiXie8 commented 2 years ago

请问你用的coco训练集,分辨率是设置的多少呢