| import torch
|
| import torch.nn as nn
|
|
|
|
|
| from feature_extraction import ViTEncoder
|
| from coarse_point_matching import CoarsePointMatching
|
| from fine_point_matching import FinePointMatching
|
| from transformer import GeometricStructureEmbedding
|
| from model_utils import sample_pts_feats
|
| from vis_utils import visualize_points_3d, features_to_colors, visualize_two_sets_3d
|
|
|
|
|
| class Net(nn.Module):
|
| def __init__(self, cfg):
|
| super(Net, self).__init__()
|
| self.cfg = cfg
|
| self.coarse_npoint = cfg.coarse_npoint
|
| self.fine_npoint = cfg.fine_npoint
|
|
|
| self.feature_extraction = ViTEncoder(cfg.feature_extraction, self.fine_npoint)
|
| self.geo_embedding = GeometricStructureEmbedding(cfg.geo_embedding)
|
| self.coarse_point_matching = CoarsePointMatching(cfg.coarse_point_matching)
|
| self.fine_point_matching = FinePointMatching(cfg.fine_point_matching)
|
|
|
| def forward(self, end_points):
|
|
|
| dense_pm, dense_fm, dense_po, dense_fo, radius = self.feature_extraction(end_points)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| bg_point = torch.ones(dense_pm.size(0),1,3).float().to(dense_pm.device) * 100
|
|
|
| sparse_pm, sparse_fm, fps_idx_m = sample_pts_feats(
|
| dense_pm, dense_fm, self.coarse_npoint, return_index=True
|
| )
|
| geo_embedding_m = self.geo_embedding(torch.cat([bg_point, sparse_pm], dim=1))
|
|
|
| sparse_po, sparse_fo, fps_idx_o = sample_pts_feats(
|
| dense_po, dense_fo, self.coarse_npoint, return_index=True
|
| )
|
| geo_embedding_o = self.geo_embedding(torch.cat([bg_point, sparse_po], dim=1))
|
|
|
|
|
|
|
| end_points, points = self.coarse_point_matching(
|
| sparse_pm, sparse_fm, geo_embedding_m,
|
| sparse_po, sparse_fo, geo_embedding_o,
|
| radius, end_points,
|
| )
|
|
|
|
|
| end_points, _ = self.fine_point_matching(
|
| dense_pm, dense_fm, geo_embedding_m, fps_idx_m,
|
| dense_po, dense_fo, geo_embedding_o, fps_idx_o,
|
| radius, end_points)
|
|
|
|
|
| return end_points, points
|
|
|
|
|
|
|
|
|
|
|
|
|
