Published on:
8 May 2024
Primary Category:
Computer Vision and Pattern Recognition
Paper Authors:
Lingdong Kong,
Xiang Xu,
Jiawei Ren,
Wenwei Zhang,
Liang Pan,
Kai Chen,
Wei Tsang Ooi,
Ziwei Liu
Integrates LiDAR and camera data without needing extra image annotations
Manipulates laser beams between scans to exploit spatial priors
Distills semantic features from images to LiDAR point clouds
Generates auxiliary labels using CLIP for unlabeled data
Achieves high accuracy with 5x fewer labels than supervised methods
Data-efficient 3D scene understanding for autonomous vehicles
This paper proposes a semi-supervised framework called LaserMix++ that leverages both LiDAR point clouds and camera images to improve 3D scene understanding for autonomous driving with far less labeled data. Key innovations include multi-modal data mixing, transferring knowledge from images to point clouds, and generating auxiliary labels from language models, which enhance regularization and feature learning.
Detecting 3D Objects from Monocular Images with LiDAR Guidance
Using lidar data to train image segmentation models
Point cloud semantic features for 3D object detection
Augmenting rare vehicles with surround-view renderings
Multi-sensor road segmentation
Dynamic neural fields for novel view LiDAR synthesis
No comments yet, be the first to start the conversation...
Sign up to comment on this paper