Distillation from visual foundation models to lidar enables frame-wise indoor semantic segmentation without manual annotations, achieving up to 56% mIoU on pseudo labels and 36% on real labels.
arXiv preprint arXiv:2104.04687 (2021)
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 5years
2026 5verdicts
UNVERDICTED 5roles
background 1polarities
background 1representative citing papers
Cross4D-JEPA uses dense projection-based cross-modal correspondence to distill features from DINOv2 or V-JEPA 2 into a 4D point encoder, outperforming intra-modal and global cross-modal baselines on four benchmarks while improving label efficiency.
HilDA pre-trains LiDAR backbones via multi-layer and global distillation from vision models plus temporal occupancy diffusion, yielding SOTA results on detection, flow, and occupancy tasks.
HGC-Det applies hyperbolic geometry to constrain cross-modal distillation between images and point clouds, with added semantic-guided voxel optimization and feature aggregation, yielding improved accuracy-efficiency trade-offs on SUN RGB-D, ARKitScenes, KITTI, and nuScenes.
The paper offers a taxonomy of 2D-to-3D adaptation strategies divided into data-centric projection, architecture-centric 3D networks, and hybrid methods that combine both.
citing papers explorer
-
Hyperbolic Distillation: Geometry-Guided Cross-Modal Transfer for Robust 3D Object Detection
HGC-Det applies hyperbolic geometry to constrain cross-modal distillation between images and point clouds, with added semantic-guided voxel optimization and feature aggregation, yielding improved accuracy-efficiency trade-offs on SUN RGB-D, ARKitScenes, KITTI, and nuScenes.