PatchContrast: Self-Supervised Pre-training for 3D Object Detection
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909577121890304 |
|---|---|
| author | Shrout, Oren Nizan, Ori Ben-Shabat, Yizhak Tal, Ayellet |
| author_facet | Shrout, Oren Nizan, Ori Ben-Shabat, Yizhak Tal, Ayellet |
| contents | Accurately detecting objects in the environment is a key challenge for autonomous vehicles. However, obtaining annotated data for detection is expensive and time-consuming. We introduce PatchContrast, a novel self-supervised point cloud pre-training framework for 3D object detection. We propose to utilize two levels of abstraction to learn discriminative representation from unlabeled data: proposal-level and patch-level. The proposal-level aims at localizing objects in relation to their surroundings, whereas the patch-level adds information about the internal connections between the object's components, hence distinguishing between different objects based on their individual components. We demonstrate how these levels can be integrated into self-supervised pre-training for various backbones to enhance the downstream 3D detection task. We show that our method outperforms existing state-of-the-art models on three commonly-used 3D detection datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2308_06985 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | PatchContrast: Self-Supervised Pre-training for 3D Object Detection Shrout, Oren Nizan, Ori Ben-Shabat, Yizhak Tal, Ayellet Computer Vision and Pattern Recognition Accurately detecting objects in the environment is a key challenge for autonomous vehicles. However, obtaining annotated data for detection is expensive and time-consuming. We introduce PatchContrast, a novel self-supervised point cloud pre-training framework for 3D object detection. We propose to utilize two levels of abstraction to learn discriminative representation from unlabeled data: proposal-level and patch-level. The proposal-level aims at localizing objects in relation to their surroundings, whereas the patch-level adds information about the internal connections between the object's components, hence distinguishing between different objects based on their individual components. We demonstrate how these levels can be integrated into self-supervised pre-training for various backbones to enhance the downstream 3D detection task. We show that our method outperforms existing state-of-the-art models on three commonly-used 3D detection datasets. |
| title | PatchContrast: Self-Supervised Pre-training for 3D Object Detection |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2308.06985 |