yolov5_tensorrt_int8 icon indicating copy to clipboard operation
yolov5_tensorrt_int8 copied to clipboard

Published 20 hours ago • verified publisher icon Wulingtian

你好为什么后处理的时候需要初始anchors?

Open henbucuoshanghai opened this issue 4 years ago • 1 comments

为什么要设置三层的参数？cur_in_tensor_params[0].nDims ？ 多谢

henbucuoshanghai avatar Apr 19 '21 03:04 henbucuoshanghai

cur_out_tensor_params是做什么的 不是很懂？ engine是已经有训练好的参数的，难道不是读取后测试图片就可以？

henbucuoshanghai avatar Apr 19 '21 03:04 henbucuoshanghai