In this paper, we propose a 3D fully convolutional encoder-decoder
archi...
Existing semantic segmentation works mainly focus on learning the contex...
Recent works have achieved great success in improving the performance of...
Attention mechanisms are widely used in salient object detection models ...