上一条： Improving Back-projection Accuracy for the Semantic Segmentation of Indoor Point Clouds with Fewer & Sparse Image Annotations

下一条： Physics-driven deep learning inversion: gradient optimization and its application to DC resistivity survey