forked from lmb-freiburg/hand3d
-
Notifications
You must be signed in to change notification settings - Fork 1
/
fine-tune.py
170 lines (149 loc) · 6.77 KB
/
fine-tune.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
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
#
# ColorHandPose3DNetwork - Network for estimating 3D Hand Pose from a single RGB Image
# Copyright (C) 2017 Christian Zimmermann
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
#
from __future__ import print_function, unicode_literals
import tensorflow as tf
import os
import sys
from nets.PosePriorNetwork import PosePriorNetwork
from data.BinaryDbReaderSTB import BinaryDbReaderSTB
from data.DomeReader import DomeReader
from utils.general import LearningRateScheduler, hand_size
# import pdb
def visualize(scoremap, hand_side, rot_mat, coord3d_can, coord3d, coord2d):
import pdb
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from utils.general import plot_hand_3d, plot_hand
import numpy as np
l = scoremap.shape[0]
for i in range(l):
l_scoremap = scoremap[i, :, :, :]
l_coord3d = coord3d[i, :, :]
l_coord3d_can = coord3d_can[i, :, :] * 6.0
l_rot_mat = rot_mat[i, :, :]
l_coord2d = coord2d[i, :, :]
if hand_side[i, 1] == 1:
print('flip')
# l_coord3d_can[:, 2] = -l_coord3d_can[:, 2]
# l_coord3d_rotate = np.dot(l_coord3d_can, l_rot_mat)
l_coord3d -= l_coord3d[0, :]
# l_coord3d_rotate *= 2.0
s = l_scoremap.shape
keypoint_coords = np.zeros((s[2], 2))
for i in range(s[2]):
v, u = np.unravel_index(np.argmax(l_scoremap[:, :, i]), (s[0], s[1]))
keypoint_coords[i, 0] = v
keypoint_coords[i, 1] = u
fig = plt.figure(1)
ax1 = fig.add_subplot(131)
ax1.imshow(np.amax(l_scoremap, axis=2))
ax2 = fig.add_subplot(132, projection='3d')
plot_hand_3d(l_coord3d, ax2, color_fixed=np.array([1.0, 0.0, 1.0]))
plot_hand_3d(l_coord3d_can, ax2, color_fixed=np.array([0.0, 1.0, 0.0]))
ax2.view_init(azim=-90.0, elev=-90.0) # aligns the 3d coord with the camera view
ax3 = fig.add_subplot(133)
plot_hand(keypoint_coords, ax3, color_fixed=np.array([0.0, 1.0, 0.0]))
plt.gca().invert_yaxis()
plt.xlabel('x')
plt.ylabel('y')
plt.show()
# Chose which variant to evaluate
# VARIANT = 'direct'
# VARIANT = 'bottleneck'
# VARIANT = 'local'
# VARIANT = 'local_w_xyz_loss'
VARIANT = 'proposed'
# training parameters
# train_para = {'lr': [1e-5, 1e-6],
# 'lr_iter': [60000],
# 'max_iter': 80000,
# 'show_loss_freq': 1000,
# 'snapshot_freq': 5000,
# 'snapshot_dir': 'snapshots_lifting_%s_dome' % VARIANT}
train_para = {'lr': [1e-4, 1e-5],
'lr_iter': [60000],
'max_iter': 120000,
'show_loss_freq': 100,
'snapshot_freq': 5000,
'org_snapshot_dir': 'snapshots_lifting_%s_my' % VARIANT,
'ft_snapshot_dir': 'snapshots_lifting_%s_my_ft' % VARIANT,
'snapshot': 'model-120000'}
# get dataset
dataset = BinaryDbReaderSTB(mode='training',
batch_size=8, shuffle=True, hand_crop=True, use_wrist_coord=False,
coord_uv_noise=True, crop_center_noise=True, crop_offset_noise=True, crop_scale_noise=True)
# build network graph
data = dataset.get()
# build network
net = PosePriorNetwork(VARIANT)
# feed trough network
evaluation = tf.placeholder_with_default(True, shape=())
_, coord3d_pred, R = net.inference(data['scoremap'], data['hand_side'], evaluation)
# cond_right = tf.equal(tf.argmax(data['hand_side'], 1), 1)
# cond_right_all = tf.tile(tf.reshape(cond_right, [-1, 1, 1]), [1, 21, 3])
# coord_xyz_can_flip = PosePriorNetwork._flip_right_hand(data['keypoint_xyz21_can'], cond_right_all)
# coord_xyz_rel_normed = tf.matmul(coord_xyz_can_flip, data['rot_mat'])
# Start TF
gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=0.4)
sess = tf.Session(config=tf.ConfigProto(gpu_options=gpu_options))
sess.run(tf.global_variables_initializer())
tf.train.start_queue_runners(sess=sess)
# Loss
loss = 0.0
if (VARIANT == 'direct') or (VARIANT == 'bottleneck'):
loss = tf.reduce_mean(tf.square(coord3d_pred - data['keypoint_xyz21_normed']))
elif VARIANT == 'local':
loss += tf.reduce_mean(tf.square(coord3d_pred - data['keypoint_xyz21_local']))
elif VARIANT == 'local_w_xyz_loss':
from utils.relative_trafo import bone_rel_trafo_inv
# transform the local coordinates back into xyz for the loss
coord3d_pred_xyz = bone_rel_trafo_inv(coord3d_pred)
loss += tf.reduce_mean(tf.square(coord3d_pred_xyz - data['keypoint_xyz21_normed']))
elif VARIANT == 'proposed':
loss += tf.reduce_mean(tf.square(coord3d_pred - data['keypoint_xyz21_can']))
loss += tf.reduce_mean(tf.square(R - data['rot_mat']))
# Solver
global_step = tf.Variable(0, trainable=False, name="global_step")
lr_scheduler = LearningRateScheduler(values=train_para['lr'], steps=train_para['lr_iter'])
lr = lr_scheduler.get_lr(global_step)
opt = tf.train.AdamOptimizer(lr)
train_op = opt.minimize(loss)
# init weights
sess.run(tf.global_variables_initializer())
saver = tf.train.Saver(max_to_keep=1, keep_checkpoint_every_n_hours=4.0)
restore = os.path.join(train_para['org_snapshot_dir'], train_para['snapshot'])
saver.restore(sess, restore)
# snapshot dir
if not os.path.exists(train_para['ft_snapshot_dir']):
os.mkdir(train_para['ft_snapshot_dir'])
print('Created snapshot dir:', train_para['ft_snapshot_dir'])
# Training loop
print('Starting to train ...')
for i in range(train_para['max_iter']):
# _, loss_v, scoremap, hand_side, rot_mat, coord3d_can, coord3d, coord2d = sess.run([train_op, loss, data['scoremap'], data['hand_side'], data['rot_mat'], coord_xyz_rel_normed, data['keypoint_xyz21'], data['keypoint_uv21']])
_, loss_v = sess.run([train_op, loss])
# visualize(scoremap, hand_side, rot_mat, coord3d_can, coord3d, coord2d)
if (i % train_para['show_loss_freq']) == 0:
print('Iteration %d\t Loss %.1e' % (i, loss_v))
sys.stdout.flush()
if (i % train_para['snapshot_freq']) == 0:
saver.save(sess, "%s/model" % train_para['ft_snapshot_dir'], global_step=i)
print('Saved a snapshot.')
sys.stdout.flush()
print('Training finished. Saving final snapshot.')
saver.save(sess, "%s/model" % train_para['ft_snapshot_dir'], global_step=train_para['max_iter'])