<ul><li>The paper introduces FACT, a method for predicting alignment quality of registered lidar point cloud pairs.</li><li>FACT extracts local features from a registered pair and processes them with a point transformer-based network to predict a misalignment class.</li><li>FACT outperforms both direct regression and prior binary classification by introducing a custom regression-by-classification loss function.</li><li>The method successfully classifies point-cloud pairs registered with ICP and GeoTransformer, and can assist in correcting misaligned point-cloud maps.</li></ul>

FACT: Multinomial Misalignment Classification for Point Cloud Registration

Discover more