ohhhyeahhh / SiamGAT

Code for the paper "Graph Attention Tracking". (CVPR2021)
https://openaccess.thecvf.com/content/CVPR2021/papers/Guo_Graph_Attention_Tracking_CVPR_2021_paper.pdf
139 stars 19 forks source link

关于TargetAware的实现 #3

Closed kongbia closed 3 years ago

kongbia commented 3 years ago

请问关于targetaware中将目标框映射到特征图大小上的 roi = torch.round((bbox + 1 - offset + stride / 2) / stride - 1).int() 这行代码可以解释为什么是这样操作吗

twotwo2 commented 3 years ago

参考iccv2015_tutorial_convolutional_feature_maps_kaiminghe和《Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition》。