Human-Image-Segmentation-with-DeepLabV3Plus-in-TensorFlow icon indicating copy to clipboard operation
Human-Image-Segmentation-with-DeepLabV3Plus-in-TensorFlow copied to clipboard
verified publisher icon nikhilroxtomar

How did you determine the layers of the backbone where the encoder should be connected

Open dubrovin-sudo opened this issue 2 years ago • 0 comments

Why we connect "conv4_block6_out" to ASPP (Atrous Spatial Pyramid Pooling) and connect "conv2_block2_out" to encoder? @nikhilroxtomar

dubrovin-sudo avatar Jan 31 '23 09:01 dubrovin-sudo