3D Feature Prediction for Masked-AutoEncoder-Based Point Cloud Pretraining
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2023
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866929328946675712 |
|---|---|
| author | Yan, Siming Yang, Yuqi Guo, Yuxiao Pan, Hao Wang, Peng-shuai Tong, Xin Liu, Yang Huang, Qixing |
| author_facet | Yan, Siming Yang, Yuqi Guo, Yuxiao Pan, Hao Wang, Peng-shuai Tong, Xin Liu, Yang Huang, Qixing |
| contents | 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 MAEs used in the image domain, where the pretext task is to restore features at the masked pixels, such as colors, the existing 3D MAE works reconstruct the missing geometry only, i.e, the location of the masked points. In contrast to previous studies, we advocate that point location recovery is inessential and restoring intrinsic point features is much superior. To this end, we propose to ignore point position reconstruction and recover high-order features at masked points including surface normals and surface variations, through a novel attention-based decoder which is independent of the encoder design. We validate the effectiveness of our pretext task and decoder design using different encoder structures for 3D training and demonstrate the advantages of our pretrained networks on various point cloud analysis tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2304_06911 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | 3D Feature Prediction for Masked-AutoEncoder-Based Point Cloud Pretraining Yan, Siming Yang, Yuqi Guo, Yuxiao Pan, Hao Wang, Peng-shuai Tong, Xin Liu, Yang Huang, Qixing Computer Vision and Pattern Recognition 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 MAEs used in the image domain, where the pretext task is to restore features at the masked pixels, such as colors, the existing 3D MAE works reconstruct the missing geometry only, i.e, the location of the masked points. In contrast to previous studies, we advocate that point location recovery is inessential and restoring intrinsic point features is much superior. To this end, we propose to ignore point position reconstruction and recover high-order features at masked points including surface normals and surface variations, through a novel attention-based decoder which is independent of the encoder design. We validate the effectiveness of our pretext task and decoder design using different encoder structures for 3D training and demonstrate the advantages of our pretrained networks on various point cloud analysis tasks. |
| title | 3D Feature Prediction for Masked-AutoEncoder-Based Point Cloud Pretraining |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2304.06911 |