当前位置:网站首页>PointNet/Pointnet++训练及测试
PointNet/Pointnet++训练及测试
2022-06-29 08:20:00 【马少爷】
1、安装Anaconda
source ~/.bashrc更新环境变量
然后可以输入conda list 测试是否成功
2、创建PyTorch环境
conda create -n myPytorch python=3.7
conda activate mypytorch
conda install pytorch1.1.0 torchvision0.3.0 -c pytorch
3、安装cloudcompare软件
snap install cloudcompare
安装meshlab软件
meshlab2020.07-linux.AppImage
4、复制pointnet项目
git clone github网址
5、完成形状分类任务
下载数据集:modelnet40_normal_resampled
解压后放在data文件夹下
python train_cls.py --model pointnet2_cls_msg --normal --log_dir pointnet2_cls_msg
使用法向量信息
若显存溢出,设置batch_size
python train_cls.py --model pointnet2_cls_msg --normal --log_dir pointnet2_cls_msg batch_size 8
对训练好的网络进行测试
python test_cls.py --normal --log_dir pointnet2_cls_msg
6、物体部件分割
使用数据集:shapenetcore_partanno_segmentation_benchmark_v0_normal
将数据集解压到data文件夹
训练命令:
python train_partseg.py --model pointnet2_part_seg_msg --normal --log_dir pointnet2_part_seg_msg
测试命令:
python test_partseg.py --normal --log_dir pointnet2_part_seg_msg
7、场景语义分割
数据集:Stanford3dDataset_v1.2_Aligned_Version
解压到:
data/s3dis/Stanford3dDataset_v1.2_Aligned_Version/
训练:
cd data_utils
python collect_indoor3d_data.py
处理后的数据保存到
data/stanford_indoor3d/
执行训练命令:
python train_semseg.py --model pointnet2_sem_seg --test_area 5 --log_dir
pointnet2_sem_seg
可视化结果保存在
log/sem_seg/pointnet2_sem_seg/visual/
执行测试命令:
python test_semseg.py --log_dir pointnet2_sem_seg --test_area 5 --visual
边栏推荐
- Tutorial on building open source Internet of things platform
- Speech signal processing - Fundamentals (I): basic acoustic knowledge
- 闭关修炼(二十五)基础web安全
- hostname -f与uname -n的返回值可能不同
- TypeScript 變量聲明 —— 類型斷言
- 递归方法 rbac菜单层级显示 无限极分类
- 对话| 数字时代,隐私计算的发展前景与挑战
- P4769-[noi2018] bubble sort [combinatorics, tree array]
- Np5 formatted output (III)
- 2022第六季完美童模 清远赛区 海选赛圆满落幕
猜你喜欢
随机推荐
Differences between x86 and x64
Voice annotation automatic segment alignment tool sppas usage notes
Official reply on issues related to the change of children's names after parents' divorce
航芯开发板&调试器
编程语言
51单片机中断与定时器计数器,基于普中科技HC6800-ESV2.0
随心玩玩(三)Mirai框架QQ机器人
华为设备配置小型网络WLAN基本业务
Target tracking [single target tracking (vot/sot), target detection, pedestrian re identification (re ID)]
Memoirs of actual combat: breaking the border from webshell
2022春夏系列 KOREANO ESSENTIAL重塑时装生命力
Operating system product key viewing method
Heavyweight released "FISCO bcos application landing guide"
ThreadLocal线程变量
实战回忆录:从Webshell开始突破边界
NP5 格式化输出(三)
The difference and usage of JS for in loop and for of loop
The sixth season of 2022 perfect children's model Qingyuan competition area audition came to a successful conclusion
Déclaration de la variable Typescript - - assertion de type
P6776-[NOI2020]超现实树







![[untitled]](/img/6e/5dd5dcff89a74f7d367c9186a77268.png)

