Hi there! I have used your implementation quite extensively already! So thank you already!
But I have a question regarding the decoder part of the network. As far as I can understand your implemetation, the only thing that is being done to the output of ASPP is trilinear interpolation. So basically, the decoder from the DeepLabv3+ paper is not implemented yet, right? Is there any reason you chose not to do it?
Hi there! I have used your implementation quite extensively already! So thank you already!
But I have a question regarding the decoder part of the network. As far as I can understand your implemetation, the only thing that is being done to the output of ASPP is trilinear interpolation. So basically, the decoder from the DeepLabv3+ paper is not implemented yet, right? Is there any reason you chose not to do it?
Thanks!