We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
还在学习中, 先记录一些关键点 , 稍后补充文字, 形成完整的笔记
tx/ty是特征图中的偏移量, 实际特征图中的每一个像素对应到原图上是多个像素, 所以特征图下的位置bx其实是个小数, 公式如下: bx = cx + sigmoid(tx), 这样保证调整偏差后bx依旧和cx是同一个格子内的点
bx = cx + sigmoid(tx)
th/tw是预设锚框缩放的比例, pw/ph是预设锚框的尺寸, 所以bw = pw * exp(tw), 这里是exp是指数运算, 具体为什么用exp我也没太弄明白, 总之就好求导, 因为指数的导数还是自己本身, 后续再深入研究明白他.
bw = pw * exp(tw)
The text was updated successfully, but these errors were encountered:
No branches or pull requests
还在学习中, 先记录一些关键点 , 稍后补充文字, 形成完整的笔记
遍历3种特征图的每一个像素, 输出tx,ty.tw,th,confidence, classes
tx/ty是特征图中的偏移量, 实际特征图中的每一个像素对应到原图上是多个像素, 所以特征图下的位置bx其实是个小数, 公式如下:
bx = cx + sigmoid(tx)
, 这样保证调整偏差后bx依旧和cx是同一个格子内的点th/tw是预设锚框缩放的比例, pw/ph是预设锚框的尺寸, 所以
bw = pw * exp(tw)
, 这里是exp是指数运算, 具体为什么用exp我也没太弄明白, 总之就好求导, 因为指数的导数还是自己本身, 后续再深入研究明白他.The text was updated successfully, but these errors were encountered: