Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2408.10007 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915295890767872 |
|---|---|
| author | Chen, Xuechao Chen, Ying Li, Jialin Nie, Qiang Deng, Hanqiu Liu, Yong Huang, Qixing Li, Yang |
| author_facet | Chen, Xuechao Chen, Ying Li, Jialin Nie, Qiang Deng, Hanqiu Liu, Yong Huang, Qixing Li, Yang |
| contents | 3D pre-training is crucial to 3D perception tasks. Nevertheless, limited by the difficulties in collecting clean and complete 3D data, 3D pre-training has persistently faced data scaling challenges. In this work, we introduce a novel self-supervised pre-training framework that incorporates millions of images into 3D pre-training corpora by leveraging a large depth estimation model. New pre-training corpora encounter new challenges in representation ability and embedding efficiency of models. Previous pre-training methods rely on farthest point sampling and k-nearest neighbors to embed a fixed number of 3D tokens. However, these approaches prove inadequate when it comes to embedding millions of samples that feature a diverse range of point numbers, spanning from 1,000 to 100,000. In contrast, we propose a tokenizer with linear-time complexity, which enables the efficient embedding of a flexible number of tokens. Accordingly, a new 3D reconstruction target is proposed to cooperate with our 3D tokenizer. Our method achieves state-of-the-art performance in 3D classification, few-shot learning, and 3D segmentation. Code is available at https://github.com/XuechaoChen/P3P-MAE. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_10007 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | P3P: Pseudo-3D Pre-training for Scaling 3D Voxel-based Masked Autoencoders Chen, Xuechao Chen, Ying Li, Jialin Nie, Qiang Deng, Hanqiu Liu, Yong Huang, Qixing Li, Yang Computer Vision and Pattern Recognition 3D pre-training is crucial to 3D perception tasks. Nevertheless, limited by the difficulties in collecting clean and complete 3D data, 3D pre-training has persistently faced data scaling challenges. In this work, we introduce a novel self-supervised pre-training framework that incorporates millions of images into 3D pre-training corpora by leveraging a large depth estimation model. New pre-training corpora encounter new challenges in representation ability and embedding efficiency of models. Previous pre-training methods rely on farthest point sampling and k-nearest neighbors to embed a fixed number of 3D tokens. However, these approaches prove inadequate when it comes to embedding millions of samples that feature a diverse range of point numbers, spanning from 1,000 to 100,000. In contrast, we propose a tokenizer with linear-time complexity, which enables the efficient embedding of a flexible number of tokens. Accordingly, a new 3D reconstruction target is proposed to cooperate with our 3D tokenizer. Our method achieves state-of-the-art performance in 3D classification, few-shot learning, and 3D segmentation. Code is available at https://github.com/XuechaoChen/P3P-MAE. |
| title | P3P: Pseudo-3D Pre-training for Scaling 3D Voxel-based Masked Autoencoders |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2408.10007 |