Papers › Distinctive 3D local deep descriptors

Distinctive 3D local deep descriptors

1 Sep 2020arXiv:2009.00258archive 2025-07-28

Fabio Poiesi, Davide Boscaini

We present a simple but yet effective method for learning distinctive 3D local deep descriptors (DIPs) that can be used to register point clouds without requiring an initial alignment. Point cloud patches are extracted, canonicalised with respect to their estimated local reference frame and encoded into rotation-invariant compact descriptors by a PointNet-based deep neural network. DIPs can effectively generalise across different sensor modalities because they are learnt end-to-end from locally and randomly sampled points. Because DIPs encode only local geometric information, they are robust to clutter, occlusions and missing regions. We evaluate and compare DIPs against alternative hand-crafted and deep descriptors on several indoor and outdoor datasets consisting of point clouds reconstructed using different sensors. Results show that DIPs (i) achieve comparable results to the state-of-the-art on RGB-D indoor scenes (3DMatch dataset), (ii) outperform state-of-the-art by a large margin on laser-scanner outdoor scenes (ETH dataset), and (iii) generalise to indoor scenes reconstructed with the Visual-SLAM system of Android ARCore. Source code: https://github.com/fabiopoiesi/dip.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

fabiopoiesi/dip officialmentioned in papermentioned on GitHubpytorch report
fabiopoiesi/gedi mentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Point Cloud Registration

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Point Cloud Registration 3DMatch Benchmark DIP Feature Matching Recall 94.8 #9 of 15 Archive leaderboard report
Point Cloud Registration ETH (trained on 3DMatch) DIP Feature Matching Recall 0.928 #3 of 20 Archive leaderboard report
Point Cloud Registration ETH (trained on 3DMatch) DIP Recall (30cm, 5 degrees) 62.41 #3 of 20 Archive leaderboard report
Point Cloud Registration FPv1 DIP RRE (degrees) 4.058 #3 of 8 Archive leaderboard report
Point Cloud Registration FPv1 DIP RTE (cm) 2.052 #3 of 8 Archive leaderboard report
Point Cloud Registration FPv1 DIP Recall (3cm, 10 degrees) 54.81 #3 of 8 Archive leaderboard report
Point Cloud Registration KITTI DIP Success Rate 97.30 #4 of 6 Archive leaderboard report
Point Cloud Registration KITTI (trained on 3DMatch) DIP Success Rate 93.51 #6 of 14 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections