Why does deconvolution loss function converge more slowly than bilinear interpolation？

PingoLH / FCHarDNet

Fully Convolutional HarDNet for Segmentation in Pytorch

MIT License

195 stars 52 forks source link

Open Tony-Hou opened 4 years ago

Tony-Hou commented 4 years ago

I use the deconv(ConvTranspose2d()) instead of bilinear, the loss is hard to descend.