Skip to main content

Table 1 Configuration of RCC-Net

From: Deep residual coalesced convolutional network for efficient semantic road segmentation

Stages Convolution # RCC
  Ordinary Asymmetric Dilated  
Input 3×480×360 n/a
Initial 19×240×180 n/a
Encoder 1 3×3 5×1;1×5 2;4;8;16;32 5
Encoder 2 3×3 5×1;1×5 2;4;8;16;32 5
Encoder 3 3×3 7×1;1×7 2;4;8;16;32 5
Decoder 1 3×3 deconvolution 3
Decoder 2 3×3 deconvolution 2
Full Conv. 11×480×360 n/a