Search Results for author:
Found 2 papers, 1 papers with code
HybridFusion: LiDAR and Vision Cross-Source Point Cloud Fusion
First, the point cloud is divided into small patches, and a matching patch set is selected based on global descriptors and spatial distribution, which constitutes the coarse matching process.
Attention Aware Cost Volume Pyramid Based Multi-view Stereo Network for 3D Reconstruction
This strategy estimates the depth map at coarsest level, while the depth maps at finer levels are considered as the upsampled depth map from previous level with pixel-wise depth residual.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
