cross-modal learning

The paper introduces a new Cross-Modal Information-Guided Network (CMIGNet) for point cloud registration. Traditional methods rely solely on features from points, leading to limitations such as lack of texture information…