VGGSfM: Visual Geometry Grounded Deep Structure From Motion

Evaluating the performance of various point cloud processing tools.

A steerer for D-dimensional keypoint descriptions is a DxD matrix that transforms the descriptions as if they were computed from a rotated image.

We write your reusable computer vision tools. 💜

PointAugment and DGCNN Classification Benchmark for TR3D

A playbook for systematically maximizing the performance of deep learning models.

Road detections from Microsoft Maps aerial imagery