DASPP
Deeper Atrous Spatial Pyramid Pooling
2000년 도입 · 논문 1편에서 사용
DASPP is a deeper version of the ASPP module (the latter from DeepLabv3) that adds standard 3 × 3 convolution after 3 × 3 dilated convolutions to refine the features and also fusing the input and the output of the DASPP module via short residual connection. Also, the number of convolution filters of ASPP is reduced from 255 to 96 to gain computational performance.
출처: LiteSeg: A Novel Lightweight ConvNet for Semantic Segmentation
소개 논문: LiteSeg: A Novel Lightweight ConvNet for Semantic Segmentation
Semantic Segmentation Modules · Computer Vision