下载资源前端资源详情
Point-PN-main.zip
大小:5.53MB
价格:10积分
下载量:0
评分:
5.0
上传者:yyfhq
更新日期:2025-09-22

wish20241105

资源文件列表(大概)

文件名
大小
datasets/
-
datasets/data_mn40.py
3.66KB
datasets/data_pn_mn40.py
3.34KB
datasets/data_pn_scan.py
7.64KB
datasets/data_scan.py
2.07KB
datasets/data_seg.py
4.29KB
LICENSE
1.04KB
logger.py
1.79KB
models/
-
models/__init__.py
162B
models/model_utils.py
1.74KB
models/point_nn.py
5.91KB
models/point_nn_seg.py
8.07KB
models/point_pn.py
8.83KB
pipeline.png
725.46KB
Point-NN_arxiv.pdf
5.25MB
pointnet2_ops_lib/
-
pointnet2_ops_lib/MANIFEST.in
29B
pointnet2_ops_lib/pointnet2_ops/
-
pointnet2_ops_lib/pointnet2_ops/__init__.py
123B
pointnet2_ops_lib/pointnet2_ops/_ext-src/
-
pointnet2_ops_lib/pointnet2_ops/_ext-src/include/
-
pointnet2_ops_lib/pointnet2_ops/_ext-src/include/ball_query.h
163B
pointnet2_ops_lib/pointnet2_ops/_ext-src/include/cuda_utils.h
1.27KB
pointnet2_ops_lib/pointnet2_ops/_ext-src/include/group_points.h
183B
pointnet2_ops_lib/pointnet2_ops/_ext-src/include/interpolate.h
386B
pointnet2_ops_lib/pointnet2_ops/_ext-src/include/sampling.h
260B
pointnet2_ops_lib/pointnet2_ops/_ext-src/include/utils.h
983B
pointnet2_ops_lib/pointnet2_ops/_ext-src/src/
-
pointnet2_ops_lib/pointnet2_ops/_ext-src/src/ball_query.cpp
1.01KB
pointnet2_ops_lib/pointnet2_ops/_ext-src/src/ball_query_gpu.cu
1.74KB
pointnet2_ops_lib/pointnet2_ops/_ext-src/src/bindings.cpp
570B
pointnet2_ops_lib/pointnet2_ops/_ext-src/src/group_points.cpp
1.91KB
pointnet2_ops_lib/pointnet2_ops/_ext-src/src/group_points_gpu.cu
2.82KB
pointnet2_ops_lib/pointnet2_ops/_ext-src/src/interpolate.cpp
3.23KB
pointnet2_ops_lib/pointnet2_ops/_ext-src/src/interpolate_gpu.cu
5.02KB
pointnet2_ops_lib/pointnet2_ops/_ext-src/src/sampling.cpp
2.83KB
pointnet2_ops_lib/pointnet2_ops/_ext-src/src/sampling_gpu.cu
6.85KB
pointnet2_ops_lib/pointnet2_ops/_version.py
22B
pointnet2_ops_lib/pointnet2_ops/pointnet2_modules.py
6.38KB
pointnet2_ops_lib/pointnet2_ops/pointnet2_utils.py
10.15KB
pointnet2_ops_lib/setup.py
1.16KB
README.md
4.82KB
requirements.txt
193B
run_nn_cls.py
4.31KB
run_nn_seg.py
5.4KB
run_pn_mn40.py
10.11KB
run_pn_scan.py
11.14KB
utils.py
12.66KB

资源内容介绍

wish20241105
# Parameter is Not All You NeedOfficial implementation of ['Parameter is Not All You Need: Starting from Non-Parametric Networks for 3D Point Cloud Analysis'](https://arxiv.org/pdf/2303.08134.pdf).The paper has been accepted by **CVPR 2023** 🔥.[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/parameter-is-not-all-you-need-starting-from/training-free-3d-point-cloud-classification)](https://paperswithcode.com/sota/training-free-3d-point-cloud-classification?p=parameter-is-not-all-you-need-starting-from)[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/parameter-is-not-all-you-need-starting-from/training-free-3d-point-cloud-classification-1)](https://paperswithcode.com/sota/training-free-3d-point-cloud-classification-1?p=parameter-is-not-all-you-need-starting-from)[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/parameter-is-not-all-you-need-starting-from/training-free-3d-part-segmentation-on)](https://paperswithcode.com/sota/training-free-3d-part-segmentation-on?p=parameter-is-not-all-you-need-starting-from)## News* **Seg-NN** has been accepted as ***CVPR 2024 Highlight Paper*** 🔥!* We release [Seg-NN](https://arxiv.org/pdf/2404.04050.pdf) and [code](https://github.com/yangyangyang127/Seg-NN), which adapts Point-NN & Point-PN into 3D scene segmentation tasks 🔥.* For the first time, we conduct 3D analysis entirely requiring $\color{darkorange}{No\ Parameter\ or\ Training\.}$ 💥* The code of Point-PN has been released 📌.* The code of Point-NN for shape classification and part segmentation has been released. ## IntroductionWe present a **N**on-parametric **N**etwork for 3D point cloud analysis, **Point-NN**, which consists of purely non-learnable components. Surprisingly, requiring no parameters or training, it performs well on various 3D tasks, and even surpasses existing fully trained models. Starting from this basic non-parametric model, we propose two extensions. First, Point-NN can serve as a base architectural framework to construct **P**arametric **N**etworks, **Point-PN**, which exhibits superior performance with simple linear layers. Second, Point-NN can be regarded as a plug-and-play module to enhance the already trained 3D models during inference by complementary knowledge.<div align="center"> <img src="pipeline.png"/></div>## Requirements### InstallationCreate a conda environment and install dependencies:```bashgit clone https://github.com/ZrrSkywalker/Point-NN.gitcd Point-NNconda create -n pointnn python=3.7conda activate pointnn# Install the according versions of torch and torchvisionconda install pytorch torchvision cudatoolkitpip install -r requirements.txtpip install pointnet2_ops_lib/.```### DatasetPlease download the following datasets: [ModelNet40](https://shapenet.cs.stanford.edu/media/modelnet40_ply_hdf5_2048.zip), [ScanObjectNN](https://hkust-vgd.ust.hk/scanobjectnn/h5_files.zip), and [ShapeNetPart](https://shapenet.cs.stanford.edu/media/shapenetcore_partanno_segmentation_benchmark_v0_normal.zip). Then, create a `data/` folder and organize the datasets as```data/|–– h5_files/|–– modelnet40_ply_hdf5_2048/|–– shapenetcore_partanno_segmentation_benchmark_v0_normal/```## Point-NN --- Very Quick Implementation 🚀### Shape ClassificationDue to the training-free manner, the preparation and inference of Point-NN only take **2 minutes**.For ModelNet40 dataset, just run:```bashpython run_nn_cls.py --dataset mn40```For ScanObjectNN dataset, just run:```bashpython run_nn_cls.py --dataset scan --split 1```Please indicate the splits at `--split` by `1,2,3` for OBJ-BG, OBJ-ONLY, and PB-T50-RS, respectively.### Part SegmentationFor ShapeNetPart, Point-NN takes **7 minutes** to achieve 71.5% mIOU (70.4% in the paper), just run:```bashpython run_nn_seg.py```You can increase the point number `--points` and k-NN neighbors `--k` into `2048` and `128`, which further acheives **74%** with 18 minutes.## Point-PN### Shape ClassificationPoint-PN is the parametric version of Point-NN with efficient parameters and simple 3D operators.For ModelNet40 dataset, just run:```bashpython run_pn_mn40.py --msg <output filename>```For ScanObjectNN dataset, just run:```bashpython run_pn_scan.py --split 1 --msg <output filename>```Please indicate the splits at `--split` by `1,2,3` for OBJ-BG, OBJ-ONLY, and PB-T50-RS, respectively.## Citation```bash@article{zhang2023parameter, title={Parameter is not all you need: Starting from non-parametric networks for 3d point cloud analysis}, author={Zhang, Renrui and Wang, Liuhui and Wang, Yali and Gao, Peng and Li, Hongsheng and Shi, Jianbo}, journal={arXiv preprint arXiv:2303.08134}, year={2023}}```## ContactIf you have any question about this project, please feel free to contact zhangrenrui@pjlab.org.cn.

用户评论 (0)

发表评论

captcha

相关资源

钢材表面缺陷检测数据集

钢材表面缺陷检测数据集是一个专门针对钢材表面缺陷检测的深度学习训练与测试资源。这个数据集的创建旨在推动钢铁工业中自动化检测技术的发展，提高生产效率和产品质量。它包含了大量经过精心标注的真实钢材图像，用于训练和验证深度学习模型，特别是针对目标检测任务的算法，如YOLO（You Only Look Once）。一、表面缺陷检测的重要性：在钢铁制造过程中，表面缺陷可能会影响材料的性能和寿命，甚至导致结构的失效。因此，及时、准确地检测出这些缺陷至关重要。传统的检测方法依赖于人工视觉检查，成本高且易受主观因素影响。随着机器学习和深度学习技术的进步，自动化检测已成为解决这一问题的有效途径。二、深度学习在表面缺陷检测中的应用：1. YOLO（You Only Look Once）：YOLO是一种实时的目标检测系统，以其快速的检测速度和相对较高的准确性而受到广泛关注。在NEU-DET数据集中，YOLO可以被训练来识别并定位钢材表面的缺陷，如裂纹、锈斑、凹痕等。2. 特征提取：深度学习模型，如卷积神经网络（CNN），能自动从图像中学习高级特征，这对于识别复杂的表面缺

6.54MB34积分

期中复习大礼包.zip

期中复习大礼包.zip

17.32MB16积分

设备配件管理和设备检修系统 SSM毕业设计 附带论文.zip

设备配件管理和设备检修系统 SSM毕业设计 附带论文启动教程:https://www.bilibili.com/video/BV1GK1iYyE2B

11.69MB39积分

程序设计课程可视化教学系统 SSM毕业设计 附带论文.zip

程序设计课程可视化教学系统 SSM毕业设计 附带论文启动教程:https://www.bilibili.com/video/BV1GK1iYyE2B

12.47MB16积分