-
Notifications
You must be signed in to change notification settings - Fork 2
/
train_shot.py
154 lines (131 loc) · 6.73 KB
/
train_shot.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
from pathlib import Path
import pytorch_lightning as pl
from torch.optim import Adam
import torch
import numpy as np
from itertools import combinations
import torch.nn.functional as F
import hydra
from dataset import ShapeNetExportDataset
import torch.nn as nn
import time
from multiprocessing import cpu_count
from pytorch_lightning import loggers as pl_loggers
from torch import optim
from pytorch_lightning.callbacks import ModelCheckpoint
from utils.util import real2prob, prob2real
class ResLayer(torch.nn.Module):
def __init__(self, dim_in, dim_out, bn=False, dropout=False) -> None:
super().__init__()
# assert(bn is False)
self.fc1 = torch.nn.Linear(dim_in, dim_out)
if bn:
self.bn1 = torch.nn.BatchNorm1d(dim_out)
else:
self.bn1 = lambda x: x
self.fc2 = torch.nn.Linear(dim_out, dim_out)
if bn:
self.bn2 = torch.nn.BatchNorm1d(dim_out)
else:
self.bn2 = lambda x: x
if dim_in != dim_out:
self.fc0 = torch.nn.Linear(dim_in, dim_out)
else:
self.fc0 = None
self.dropout = nn.Dropout(0.2) if dropout else nn.Identity()
def forward(self, x):
x_res = x if self.fc0 is None else self.fc0(x)
x = F.relu(self.bn1(self.fc1(x)))
x = self.bn2(self.fc2(x))
return self.dropout(x + x_res)
class BeyondCPPF(pl.LightningModule):
def __init__(self, cfg) -> None:
super().__init__()
self.cfg = cfg
# self.vis = Visdom(port=12345)
fcs_shot = [352,] + [128] * 5 + [64,]
self.shot_encoder = nn.Sequential(
*[ResLayer(fcs_shot[i], fcs_shot[i + 1], False) for i in range(len(fcs_shot) - 1)]
)
input_dim = len(list(combinations(np.arange(cfg.num_more + 2), 2))) * 4 + (cfg.num_more + 2) * 64
output_dim = 256 # 3 for scale
fcs = [input_dim] + [128] * 5 + [output_dim]
self.tuple_encoder = nn.Sequential(
*[ResLayer(fcs[i], fcs[i + 1], False) for i in range(len(fcs) - 1)]
)
self.logit_encoder = nn.Sequential(
ResLayer(256, 256, False),
ResLayer(256, 256, False),
ResLayer(256, 64 * 3, False),
)
self.scale_encoder = nn.Sequential(
ResLayer(256, 128, False),
ResLayer(128, 64, False),
ResLayer(64, 3, False),
)
def prepare_tuple_inputs(self, points, point_idxs_all, shot_feat, normal):
shot_inputs = torch.cat([shot_feat[point_idxs_all[:, i]] for i in range(0, self.cfg.num_more + 2)], -1)
normal_inputs = torch.cat([torch.max(torch.sum(normal[point_idxs_all[:, i]] * normal[point_idxs_all[:, j]], dim=-1, keepdim=True),
torch.sum(-normal[point_idxs_all[:, i]] * normal[point_idxs_all[:, j]], dim=-1, keepdim=True))
for (i, j) in combinations(np.arange(point_idxs_all.shape[-1]), 2)], -1)
coord_inputs = torch.cat([points[point_idxs_all[:, i]] - points[point_idxs_all[:, j]] for (i, j) in combinations(np.arange(point_idxs_all.shape[-1]), 2)], -1)
inputs = torch.cat([coord_inputs, normal_inputs, shot_inputs], -1)
return inputs
def training_step(self, batch, batch_idx):
pc_canon = batch['pc_canon'][0]
points = batch['pc'][0]
point_idxs_all = torch.from_numpy(np.random.randint(0, points.shape[0], (10000, self.cfg.num_more + 2))).long().cuda()
shot_feat = self.shot_encoder(batch['shot'][0])
# shot_feat.fill_(0)
normal = batch['normal'][0]
inputs = self.prepare_tuple_inputs(points, point_idxs_all, shot_feat, normal)
# pca_quat = matrix_to_quaternion(pca_basis)
feat = self.tuple_encoder(inputs)
preds_cls = self.logit_encoder(feat).reshape(feat.shape[0], 6, -1)
with torch.no_grad():
target_cls = real2prob(torch.clamp(pc_canon[point_idxs_all[:, :2]], -0.5, 0.5) + 0.5, 1., preds_cls.shape[-1]).reshape(feat.shape[0], 6, -1) # 2NP x 3 x NBIN
loss_cls = F.kl_div(F.log_softmax(preds_cls, dim=-1), target_cls, reduction='batchmean')
preds_scale = self.scale_encoder(feat)
target_scale = batch['bound'][0]
loss_scale = F.mse_loss(preds_scale, target_scale[None].expand_as(preds_scale))
loss = loss_cls + loss_scale
# self.log('loss', loss, prog_bar=True)
self.log('cls', loss_cls, prog_bar=True)
self.log('scale', loss_scale, prog_bar=True)
# vis_weights = torch_scatter.scatter_add(preds_weight[:, 1:2].expand_as(point_idxs_all).reshape(-1), point_idxs_all.reshape(-1), dim_size=points.shape[0]).detach()
# vis_weights /= vis_weights.max()
# if batch_idx % 10 == 0:
# # print(vis_weights.max(), vis_weights.mean(), (vis_weights > 0).sum())
# cmap = cm.get_cmap('jet')
# self.vis.scatter(points.cpu().numpy(), win=4, opts=dict(markersize=3, markercolor=(np.array([cmap(w)[:3] for w in vis_weights.cpu().numpy()]) * 255.).astype(np.uint8)))
return loss
def forward(self, points, point_idxs_all, shot_feat, normal):
inputs = self.prepare_tuple_inputs(points, point_idxs_all, self.shot_encoder(shot_feat), normal)
feat = self.tuple_encoder(inputs)
preds_scale = self.scale_encoder(feat)
preds_cls = self.logit_encoder(feat).reshape(feat.shape[0], 6, -1)
return preds_cls, preds_scale
def configure_optimizers(self):
opt = Adam([*self.shot_encoder.parameters(),
*self.tuple_encoder.parameters(),
*self.logit_encoder.parameters(),
*self.scale_encoder.parameters()],
lr=self.cfg.opt.lr, weight_decay=self.cfg.opt.weight_decay)
return [opt], [optim.lr_scheduler.StepLR(opt, 25, 0.5)]
import wandb
@hydra.main(config_path='./config', config_name='config', version_base='1.2')
def train(cfg):
hydra_cfg = hydra.core.hydra_config.HydraConfig.get()
output_dir = hydra_cfg['runtime']['output_dir']
model_ckpt = ModelCheckpoint(save_last=True, save_top_k=-1, every_n_epochs=10, filename='{epoch}')
pl_module = BeyondCPPF(cfg)
trainer = pl.Trainer(max_epochs=101, accelerator='auto', callbacks=[model_ckpt],
logger=pl_loggers.TensorBoardLogger(save_dir=output_dir),
detect_anomaly=False) # check_val_every_n_epoch=10)
def init_fn(i):
return np.random.seed(round(time.time() * 1000) % (2 ** 32) + i)
ds = ShapeNetExportDataset(cfg)
df = torch.utils.data.DataLoader(ds, pin_memory=True, batch_size=1, shuffle=True, num_workers=cpu_count() // 2, worker_init_fn=init_fn)
trainer.fit(pl_module, df)
if __name__ == '__main__':
train()