Multi-View Representation is What You Need for Point-Cloud Pre-Training

[PDF] arxiv.org

3d feature prediction for masked-autoencoder-based point cloud pretraining

S Yan, Y Yang, Y Guo, H Pan, P Wang, X Tong… - arXiv preprint arXiv …, 2023 - arxiv.org
Masked autoencoders (MAE) have recently been introduced to 3D self-supervised
pretraining for point clouds due to their great success in NLP and computer vision. Unlike …
被引用次数：11 相关文章 所有 3 个版本
[PDF] arxiv.org

P3P: Pseudo-3D Pre-training for Scaling 3D Masked Autoencoders

X Chen, Y Chen, J Li, Q Nie, Y Liu, Q Huang… - arXiv preprint arXiv …, 2024 - arxiv.org
3D pre-training is crucial to 3D perception tasks. However, limited by the difficulties in
collecting clean 3D data, 3D pre-training consistently faced data scaling challenges …
相关文章 所有 2 个版本