基于ROPNet项目训练modelnet40数据集进行3d点云的配置

项目地址： https://github.com/zhulf0804/ROPNet 在 MVP Registration Challenge (ICCV Workshop 2021)（ICCV Workshop 2021）中获得了第二名。项目可以在win10环境下运行。
论文地址： https://arxiv.org/abs/2107.02583

网络简介： 一种新的深度学习模型，该模型利用具有区别特征的代表性重叠点进行配准，将部分到部分配准转换为部分完全配准。基于pointnet输出的特征设计了一个上下文引导模块，使用一个编码器来提取全局特征来预测点重叠得分。为了更好地找到有代表性的重叠点，使用提取的全局特征进行粗对齐。然后，引入一种变压器来丰富点特征，并基于点重叠得分和特征匹配去除非代表性点。在部分到完全的模式下建立相似度矩阵，最后采用加权支持向量差来估计变换矩阵。
在这里插入图片描述
实施效果: 从数据上看ROPNet与RPMNet与保持了断崖式的领先地位

1、运行环境安装

1.1 项目下载

打开https://github.com/zhulf0804/ROPNet，点Download ZIP然后将代码解压到指定目录下即可。
在这里插入图片描述

1.2 依赖项安装

在装有pytorch的环境终端，进入ROPNet-master/src目录，执行以下安装命令。如果已经安装了torch 环境和open3d包，则不用再进行安装了

pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118pip install open3d

1.3 模型与数据下载

modelnet40数据集 here [435M]
数据集下载后存储为以下路径即可。
在这里插入图片描述

官网预训练模型，无。
第三方预训练模型：使用ROPNet项目在modelnet40数据集上训练的模型

2、关键代码

2.1 dataloader

作者所提供的dataloader只能加载https://shapenet.cs.stanford.edu/media/modelnet40_ply_hdf5_2048.zip 数据集，其所返回的tgt_cloud, src_cloud实质上是基于一个点云采样而来的。 其中的self.label2cat, self.cat2label， self.symmetric_labels等对象代码实际上是没有任何作用的。

import copy
import h5py
import math
import numpy as np
import os
import torchfrom torch.utils.data import Dataset
import sysBASE_DIR = os.path.dirname(os.path.abspath(__file__))
ROOR_DIR = os.path.dirname(BASE_DIR)
sys.path.append(ROOR_DIR)
from utils import  random_select_points, shift_point_cloud, jitter_point_cloud, \generate_random_rotation_matrix, generate_random_tranlation_vector, \transform, random_crop, shuffle_pc, random_scale_point_cloud, flip_pchalf1 = ['airplane', 'bathtub', 'bed', 'bench', 'bookshelf', 'bottle', 'bowl','car', 'chair', 'cone', 'cup', 'curtain', 'desk', 'door', 'dresser','flower_pot', 'glass_box', 'guitar', 'keyboard', 'lamp']
half1_symmetric = ['bottle', 'bowl', 'cone', 'cup', 'flower_pot', 'lamp']half2 = ['laptop', 'mantel', 'monitor', 'night_stand', 'person', 'piano','plant', 'radio', 'range_hood', 'sink', 'sofa', 'stairs', 'stool','table', 'tent', 'toilet', 'tv_stand', 'vase', 'wardrobe', 'xbox']
half2_symmetric = ['tent', 'vase']class ModelNet40(Dataset):def __init__(self, root, split, npts, p_keep, noise, unseen, ao=False,normal=False):super(ModelNet40, self).__init__()self.single = False # for specific-class visualizationassert split in ['train', 'val', 'test']self.split = splitself.npts = nptsself.p_keep = p_keepself.noise = noiseself.unseen = unseenself.ao = ao # Asymmetric Objectsself.normal = normalself.half = half1 if split in 'train' else half2self.symmetric = half1_symmetric + half2_symmetricself.label2cat, self.cat2label = self.label2category(os.path.join(root, 'shape_names.txt'))self.half_labels = [self.cat2label[cat] for cat in self.half]self.symmetric_labels = [self.cat2label[cat] for cat in self.symmetric]files = [os.path.join(root, 'ply_data_train{}.h5'.format(i))for i in range(5)]if split == 'test':files = [os.path.join(root, 'ply_data_test{}.h5'.format(i))for i in range(2)]self.data, self.labels = self.decode_h5(files)print(f'split: {self.split}, unique_ids: {len(np.unique(self.labels))}')if self.split == 'train':self.Rs = [generate_random_rotation_matrix() for _ in range(len(self.data))]self.ts = [generate_random_tranlation_vector() for _ in range(len(self.data))]def label2category(self, file):with open(file, 'r') as f:label2cat = [category.strip() for category in f.readlines()]cat2label = {label2cat[i]: i for i in range(len(label2cat))}return label2cat, cat2labeldef decode_h5(self, files):points, normal, label = [], [], []for file in files:f = h5py.File(file, 'r')cur_points = f['data'][:].astype(np.float32)cur_normal = f['normal'][:].astype(np.float32)cur_label = f['label'][:].flatten().astype(np.int32)if self.unseen:idx = np.isin(cur_label, self.half_labels)cur_points = cur_points[idx]cur_normal = cur_normal[idx]cur_label = cur_label[idx]if self.ao and self.split in ['val', 'test']:idx = ~np.isin(cur_label, self.symmetric_labels)cur_points = cur_points[idx]cur_normal = cur_normal[idx]cur_label = cur_label[idx]if self.single:idx = np.isin(cur_label, [8])cur_points = cur_points[idx]cur_normal = cur_normal[idx]cur_label = cur_label[idx]points.append(cur_points)normal.append(cur_normal)label.append(cur_label)points = np.concatenate(points, axis=0)normal = np.concatenate(normal, axis=0)data = np.concatenate([points, normal], axis=-1).astype(np.float32)label = np.concatenate(label, axis=0)return data, labeldef compose(self, item, p_keep):tgt_cloud = self.data[item, ...]if self.split != 'train':np.random.seed(item)R, t = generate_random_rotation_matrix(), generate_random_tranlation_vector()else:tgt_cloud = flip_pc(tgt_cloud)R, t = generate_random_rotation_matrix(), generate_random_tranlation_vector()src_cloud = random_crop(copy.deepcopy(tgt_cloud), p_keep=p_keep[0])src_size = math.ceil(self.npts * p_keep[0])tgt_size = self.nptsif len(p_keep) > 1:tgt_cloud = random_crop(copy.deepcopy(tgt_cloud),p_keep=p_keep[1])tgt_size = math.ceil(self.npts * p_keep[1])src_cloud_points = transform(src_cloud[:, :3], R, t)src_cloud_normal = transform(src_cloud[:, 3:], R)src_cloud = np.concatenate([src_cloud_points, src_cloud_normal],axis=-1)src_cloud = random_select_points(src_cloud, m=src_size)tgt_cloud = random_select_points(tgt_cloud, m=tgt_size)if self.split == 'train' or self.noise:src_cloud[:, :3] = jitter_point_cloud(src_cloud[:, :3])tgt_cloud[:, :3] = jitter_point_cloud(tgt_cloud[:, :3])tgt_cloud, src_cloud = shuffle_pc(tgt_cloud), shuffle_pc(src_cloud)return src_cloud, tgt_cloud, R, tdef __getitem__(self, item):src_cloud, tgt_cloud, R, t = self.compose(item=item,p_keep=self.p_keep)if not self.normal:tgt_cloud, src_cloud = tgt_cloud[:, :3], src_cloud[:, :3]return tgt_cloud, src_cloud, R, tdef __len__(self):return len(self.data)

2.2 模型设计

模型设计如下：
在这里插入图片描述

2.3 loss设计

其主要包含Init_loss、Refine_loss和Ol_loss。
其中Init_loss是用于计算 $预测点云_0$ 与目标点云的mse或mae loss，
Refine_loss用于计算 $预测点云_{[1:]}$ 与目标点云的加权mae loss
Ol_loss用于计算两个输入点云输出的重叠分数，使两个点云对应点的重叠分数是一样的。
在这里插入图片描述

具体实现代码如上：


import math
import torch
import torch.nn as nn
from utils import square_distsdef Init_loss(gt_transformed_src, pred_transformed_src, loss_type='mae'):losses = {}num_iter = 1if loss_type == 'mse':criterion = nn.MSELoss(reduction='mean')for i in range(num_iter):losses['mse_{}'.format(i)] = criterion(pred_transformed_src[i],gt_transformed_src)elif loss_type == 'mae':criterion = nn.L1Loss(reduction='mean')for i in range(num_iter):losses['mae_{}'.format(i)] = criterion(pred_transformed_src[i],gt_transformed_src)else:raise NotImplementedErrortotal_losses = []for k in losses:total_losses.append(losses[k])losses = torch.sum(torch.stack(total_losses), dim=0)return lossesdef Refine_loss(gt_transformed_src, pred_transformed_src, weights=None, loss_type='mae'):losses = {}num_iter = len(pred_transformed_src)for i in range(num_iter):if weights is None:losses['mae_{}'.format(i)] = torch.mean(torch.abs(pred_transformed_src[i] - gt_transformed_src))else:losses['mae_{}'.format(i)] = torch.mean(torch.sum(weights * torch.mean(torch.abs(pred_transformed_src[i] -gt_transformed_src), dim=-1)/ (torch.sum(weights, dim=-1, keepdim=True) + 1e-8), dim=-1))total_losses = []for k in losses:total_losses.append(losses[k])losses = torch.sum(torch.stack(total_losses), dim=0)return lossesdef Ol_loss(x_ol, y_ol, dists):CELoss = nn.CrossEntropyLoss()x_ol_gt = (torch.min(dists, dim=-1)[0] < 0.05 * 0.05).long() # (B, N)y_ol_gt = (torch.min(dists, dim=1)[0] < 0.05 * 0.05).long() # (B, M)x_ol_loss = CELoss(x_ol, x_ol_gt)y_ol_loss = CELoss(y_ol, y_ol_gt)ol_loss = (x_ol_loss + y_ol_loss) / 2return ol_lossdef cal_loss(gt_transformed_src, pred_transformed_src, dists, x_ol, y_ol):losses = {}losses['init'] = Init_loss(gt_transformed_src,pred_transformed_src[0:1])if x_ol is not None:losses['ol'] = Ol_loss(x_ol, y_ol, dists)losses['refine'] = Refine_loss(gt_transformed_src,pred_transformed_src[1:],weights=None)alpha, beta, gamma = 1, 0.1, 1if x_ol is not None:losses['total'] = losses['init'] + beta * losses['ol'] + gamma * losses['refine']else:losses['total'] = losses['init'] + losses['refine']return losses

3、训练与预测

先进入src目录，并将modelnet40_ply_hdf5_2048.zip解压在src目录下
在这里插入图片描述

3.1 训练

训练命令及训练输出如下所示

python train.py --root modelnet40_ply_hdf5_2048/ --noise --unseen

python 请添加图片描述
在训练过程中会在work_dirs\models\checkpoints目录下生成两个模型文件
在这里插入图片描述

3.2 验证

训练命令及训练输出如下所示

python eval.py --root modelnet40_ply_hdf5_2048/  --unseen --noise  --cuda --checkpoint work_dirs/models/checkpoints/min_rot_error.pth

请添加图片描述

3.3 测试

测试训练数据的命令如下

python vis.py --root modelnet40_ply_hdf5_2048/  --unseen --noise  --checkpoint work_dirs/models/checkpoints/min_rot_error.pth

具体配准效果如下所示，其中绿色点云为输入点云，红色点云为参考点云，蓝色点云为配准后的点云。可以看到蓝色点云基本与红色点云重合，可以确定其配准效果十分完好。
在这里插入图片描述

3.4 处理自己的数据集

基于该项目训练并处理自己数据的教程后续会给出。