 The proposed Lightweight Dynamic Edition Network, LBANET, is a novel approach for automatic road extraction from remote sensing images. It uses an improved asymmetric convolution block, ACB, based inception structure to extend the low-level features in the feature extraction layer. Additionally, it introduces depth-wise separable convolution, DSC, to reduce the computational complexity of the model and an adaptation-weighted overlay to capture the salient features. Furthermore, a dynamic weighted combined loss is used to solve the sample imbalance problem and improve segmentation accuracy. The results show that LBANET outperforms other state-of-the-art methods in terms of accuracy and runtime efficiency. This article was authored by Bohoalu, Jin Lee Ding, GSO, and others.