391 lines
17 KiB
Python
391 lines
17 KiB
Python
import os
|
|
import sys
|
|
from configs.iccv.config_cvae_curv import config
|
|
|
|
os.environ["CUDA_VISIBLE_DEVICES"] = config['gpu']
|
|
import tensorflow as tf
|
|
|
|
gpus = tf.config.experimental.list_physical_devices('GPU')
|
|
if gpus:
|
|
try:
|
|
# Currently, memory growth needs to be the same across GPUs
|
|
for gpu in gpus:
|
|
tf.config.experimental.set_memory_growth(gpu, True)
|
|
logical_gpus = tf.config.experimental.list_logical_devices('GPU')
|
|
print(len(gpus), "Physical GPUs,", len(logical_gpus), "Logical GPUs")
|
|
except RuntimeError as e:
|
|
# Memory growth must be set before GPUs have been initialized
|
|
print(e)
|
|
|
|
import time
|
|
import argparse
|
|
import wandb
|
|
import random
|
|
import cmapy as cm
|
|
import numpy as np
|
|
import scipy as sp
|
|
from custom_utils.custom_utils import UV2Mesh, UV2MeshLabel, maskFacesBatch, _maskFaces, load_model, get_summary_html, get_summary_str
|
|
import plotly.graph_objects as go
|
|
import plotly.express as ex
|
|
from plotly.subplots import make_subplots
|
|
from PIL import Image
|
|
|
|
# from losses import collision_penalty as collision_penalty
|
|
from losses import collision_loss as collision_penalty
|
|
from dataset import DatasetFactory, StaticDataset
|
|
from custom_utils.custom_utils import batch_to_dict
|
|
from custom_utils.smpl import SMPL
|
|
from models.cvae_curv_model import CVAE
|
|
from losses import kl_loss
|
|
from custom_utils.custom_utils import Accumulator
|
|
|
|
|
|
|
|
def init_wandb():
|
|
run = wandb.init(project='iccv', name=config['name'], notes=config['notes'], dir='/tmp', tags=config['tags'])
|
|
|
|
artifact = wandb.Artifact('code', type='code')
|
|
artifact.add_file('cvae_curv.py')
|
|
artifact.add_file('models/cvae_curv_model.py')
|
|
artifact.add_file('models/cvae_model.py')
|
|
artifact.add_file('dataset.py')
|
|
artifact.add_file('losses.py')
|
|
artifact.add_file('models/base.py')
|
|
artifact.add_file('models/cond_encoder.py')
|
|
artifact.add_file('dataset.py')
|
|
artifact.add_file('convnext/models/convnext_tf.py')
|
|
config_file_names = [key for key in sys.modules.keys() if key.startswith('configs.')]
|
|
for config_file_name in config_file_names:
|
|
file_name = '/'.join(config_file_name.split('.')) + '.py'
|
|
if os.path.exists(file_name):
|
|
artifact.add_file(file_name)
|
|
run.log_artifact(artifact)
|
|
|
|
uv2mesh = UV2MeshLabel(res=256)
|
|
|
|
|
|
def plot_3d(curv_gt, curv_pred, faces, mesh):
|
|
colors = np.array([cm.color('jet', 1 - i) for i in np.linspace(0, 1, 6+1)])
|
|
v_color_1 = np.zeros((mesh.shape[0], 3))
|
|
v_color_2 = np.zeros((mesh.shape[0], 3))
|
|
for i in range(6+1):
|
|
v_color_1[curv_gt == i, :] = colors[i]
|
|
v_color_2[curv_pred == i, :] = colors[i]
|
|
|
|
fig = make_subplots(rows=1, cols=2, start_cell="top-left", specs = [ [{ 'type': 'scene' } for _ in range(2)] ])
|
|
|
|
fig.add_trace(go.Mesh3d(
|
|
# 8 vertices of a cube
|
|
x=mesh[:,0],
|
|
y=mesh[:,1],
|
|
z=mesh[:,2],
|
|
# i, j and k give the vertices of triangles
|
|
i = faces[:,0],
|
|
j = faces[:,1],
|
|
k = faces[:,2],
|
|
vertexcolor=v_color_1,
|
|
text='gt'
|
|
),
|
|
row=1, col=1)
|
|
|
|
fig.add_trace(go.Mesh3d(
|
|
# 8 vertices of a cube
|
|
x=mesh[:,0],
|
|
y=mesh[:,1],
|
|
z=mesh[:,2],
|
|
# i, j and k give the vertices of triangles
|
|
i = faces[:,0],
|
|
j = faces[:,1],
|
|
k = faces[:,2],
|
|
vertexcolor=v_color_2,
|
|
text='pred'
|
|
),
|
|
row=1, col=2)
|
|
|
|
|
|
fig.update_layout(scene_aspectmode='manual', scene_aspectratio=dict(x=1, y=1, z=1),
|
|
scene2_aspectmode='manual', scene2_aspectratio=dict(x=1, y=1, z=1),
|
|
)
|
|
helper = dict(xaxis = dict(nticks=4, range=[-1,1],),
|
|
yaxis = dict(nticks=4, range=[-1,1],),
|
|
zaxis = dict(nticks=4, range=[-1,1],),)
|
|
fig.update_layout(scene = helper,
|
|
scene2 = helper,
|
|
)
|
|
|
|
return fig
|
|
|
|
|
|
|
|
g_mean_iou = tf.keras.metrics.MeanIoU(num_classes=7)
|
|
def get_mean_iou(y_true, y_pred):
|
|
g_mean_iou.reset_state()
|
|
g_mean_iou.update_state(y_true, y_pred)
|
|
return g_mean_iou.result()
|
|
|
|
# https://github.com/NifTK/NiftyNet/blob/dev/niftynet/layer/loss_segmentation.py
|
|
def dice_dense(prediction, ground_truth, weight_map=None):
|
|
"""
|
|
Computing mean-class Dice similarity.
|
|
:param prediction: last dimension should have ``num_classes``
|
|
:param ground_truth: segmentation ground truth (encoded as a binary matrix)
|
|
last dimension should be ``num_classes``
|
|
:param weight_map:
|
|
:return: ``1.0 - mean(Dice similarity per class)``
|
|
"""
|
|
|
|
if weight_map is not None:
|
|
raise NotImplementedError
|
|
prediction = tf.cast(prediction, dtype=tf.float32)
|
|
ground_truth = tf.cast(ground_truth, dtype=tf.float32)
|
|
ground_truth = tf.reshape(ground_truth, prediction.shape)
|
|
# computing Dice over the spatial dimensions
|
|
reduce_axes = list(range(len(prediction.shape) - 1))
|
|
dice_numerator = 2.0 * tf.reduce_sum(
|
|
prediction * ground_truth, axis=reduce_axes)
|
|
dice_denominator = \
|
|
tf.reduce_sum(tf.square(prediction), axis=reduce_axes) + \
|
|
tf.reduce_sum(tf.square(ground_truth), axis=reduce_axes)
|
|
|
|
epsilon = 0.00001
|
|
|
|
dice_score = (dice_numerator + epsilon) / (dice_denominator + epsilon)
|
|
return tf.reduce_mean(dice_score)
|
|
|
|
|
|
@tf.function
|
|
def train_step(batch):
|
|
with tf.GradientTape() as tape:
|
|
|
|
pred, mean, logvar = model([batch['uv_static'], batch['uv_body_posed'], batch['uv_curvature'], batch['uv_mask']], training=True)
|
|
|
|
ce_uv_loss = scce(tf.math.argmax(batch['uv_curvature'], axis=-1), pred)
|
|
dice = dice_dense(pred, batch['uv_curvature'])
|
|
|
|
mesh_pred = uv2mesh(pred) * batch['mesh_mask'][:,:,tf.newaxis]
|
|
mesh_orig = tf.cast(uv2mesh(tf.math.argmax(batch['uv_curvature'], axis=-1, output_type=tf.int32)), tf.float32) * batch['mesh_mask']
|
|
ce_mesh_loss = scce(mesh_orig, mesh_pred)
|
|
|
|
reconstr_loss = ce_uv_loss + ce_mesh_loss
|
|
# reconstr_loss = 1 - dice
|
|
reconstr_loss = tf.reduce_mean(reconstr_loss)
|
|
kl = kl_loss(mean, logvar)
|
|
|
|
loss = reconstr_loss + g_balancing_coeff * kl
|
|
mean = tf.reduce_mean(mean)
|
|
logvar = tf.reduce_mean(logvar)
|
|
|
|
|
|
mean_iou = get_mean_iou(tf.math.argmax(batch['uv_curvature'], axis=-1), tf.math.argmax(pred, axis=-1))
|
|
|
|
|
|
gradients = tape.gradient(loss, model.trainable_variables)
|
|
optimizer.apply_gradients(zip(gradients, model.trainable_variables))
|
|
return { 'train/loss': loss, 'train/ce_uv_loss': ce_uv_loss, 'train/mean_iou': mean_iou, 'train/dice': dice,
|
|
'train/reconstr_loss': reconstr_loss, 'train/ce_mesh_loss': ce_mesh_loss,
|
|
'train/mean': mean, 'train/logvar': logvar, 'train/kl_loss': kl,'train/var': np.e ** logvar,
|
|
}
|
|
|
|
|
|
@tf.function
|
|
def valid_step(batch):
|
|
pred, mean, logvar = model([batch['uv_static'], batch['uv_body_posed'], batch['uv_curvature'], batch['uv_mask']], training=False)
|
|
|
|
ce_uv_loss = scce(tf.math.argmax(batch['uv_curvature'], axis=-1), pred)
|
|
loss = ce_uv_loss
|
|
|
|
mean_iou = get_mean_iou(tf.math.argmax(batch['uv_curvature'], axis=-1), tf.math.argmax(pred, axis=-1))
|
|
dice = dice_dense(pred, batch['uv_curvature'])
|
|
|
|
reconstr_loss = ce_uv_loss
|
|
# reconstr_loss = 1 - dice
|
|
reconstr_loss = tf.reduce_mean(reconstr_loss)
|
|
kl = kl_loss(mean, logvar)
|
|
|
|
loss = reconstr_loss + g_balancing_coeff * kl
|
|
mean = tf.reduce_mean(mean)
|
|
logvar = tf.reduce_mean(logvar)
|
|
|
|
mesh_pred = uv2mesh(pred) * batch['mesh_mask'][:,:,tf.newaxis]
|
|
mesh_orig = tf.cast(uv2mesh(tf.math.argmax(batch['uv_curvature'], axis=-1, output_type=tf.int32)), tf.float32) * batch['mesh_mask']
|
|
ce_mesh_loss = scce(mesh_orig, mesh_pred)
|
|
|
|
return { 'valid/loss': loss, 'valid/ce_uv_loss': ce_uv_loss, 'valid/mean_iou': mean_iou, 'valid/dice': dice,
|
|
'valid/reconstr_loss': reconstr_loss, 'valid/ce_mesh_loss': ce_mesh_loss,
|
|
'valid/mean': mean, 'valid/logvar': logvar, 'valid/kl_loss': kl, 'valid/var': np.e ** logvar }
|
|
|
|
|
|
def plot_everything(batch: dict, type: str):
|
|
pred, mean, logvar = model([batch['uv_static'], batch['uv_body_posed'], batch['uv_curvature'], batch['uv_mask']], training=False)
|
|
|
|
pred = np.argmax(pred.numpy(), axis=-1)
|
|
mesh_pred = uv2mesh(pred)
|
|
mesh_orig = uv2mesh(tf.math.argmax(batch['uv_curvature'], axis=-1)) # batch['mesh_curvature']
|
|
mesh_base = batch['mesh_posed']
|
|
|
|
orig_img = batch['uv_curvature']
|
|
pred_img = pred
|
|
img = tf.concat((tf.math.argmax(orig_img, axis=-1), pred_img), axis=2)
|
|
|
|
colors = np.array([cm.color('jet', 1 - i) for i in np.linspace(0, 1, 6+1)])
|
|
color_img = np.zeros((img.shape[0], img.shape[1], img.shape[2], 3), np.uint8)
|
|
for i in range(7):
|
|
color_img[img == i, :] = colors[i]
|
|
|
|
imgs = [wandb.Image(img[i].numpy().astype(np.float32) / 6.0) for i in range(4)]
|
|
color_img = [wandb.Image(color_img[i]) for i in range(4)]
|
|
|
|
F = np.load("custom_utils/smpl_faces.npy")
|
|
mask = batch['mesh_mask'].numpy().astype(bool)
|
|
_, faces = maskFacesBatch(np.zeros((4, 14475, 3)), F, mask)
|
|
gt_curv = [mesh_orig[i].numpy()[mask[i]] for i in range(pred.shape[0])]
|
|
pred_curv = [mesh_pred[i].numpy()[mask[i]] for i in range(pred.shape[0])]
|
|
mesh = [mesh_base[i].numpy()[mask[i]] for i in range(pred.shape[0])]
|
|
|
|
figs = { type + "/garment_" + str(i): plot_3d(gt_curv[i], pred_curv[i], faces[i], mesh[i]) for i in range(4) }
|
|
|
|
mean_hist = wandb.Histogram(sequence=mean.numpy().mean(axis=0), num_bins=model.latent_dim)
|
|
logvar_hist = wandb.Histogram(sequence=logvar.numpy().mean(axis=0), num_bins=model.latent_dim)
|
|
|
|
return { **figs, **{ type + '/images': imgs, type + '/images_color': color_img, type + '/mean_hist': mean_hist, type + '/logvar_hist': logvar_hist } }
|
|
|
|
|
|
|
|
def sample():
|
|
iteration = load_model(model, f'checkpoints/iccv/cvae_curv_3', return_iteration=True)
|
|
for batch in dataset_train:
|
|
batch = static_dataset_train.expand_batch(batch_to_dict(batch, config['signature_dynamic']))
|
|
nr_of_samples = 4
|
|
|
|
for i in range(4):
|
|
# pred = model.sample(batch['uv_static'], batch['uv_body_posed'], batch['uv_mask'])
|
|
# pred = model.sample_single(batch['uv_static'][0], batch['uv_body_posed'][0], batch['uv_mask'][0], nr_of_samples=nr_of_samples)
|
|
pred = model.sample(tf.tile(batch['uv_static'][i][None,:,:,:], [nr_of_samples,1,1,1]), tf.tile(batch['uv_body_posed'][i][None,:,:,:], [nr_of_samples,1,1,1]), tf.tile(batch['uv_mask'][i][None,:,:], [nr_of_samples,1,1]))
|
|
# pred,_,_ = model([tf.tile(batch['uv_static'][0][None,:,:,:], [nr_of_samples,1,1,1]), tf.tile(batch['uv_body_posed'][0][None,:,:,:], [nr_of_samples,1,1,1]), tf.tile(batch['uv_mask'][0][None,:,:], [nr_of_samples,1,1])], training=False)
|
|
# pred = model.call_sample([batch['uv_static'], batch['uv_body_posed'], batch['uv_curvature'], batch['uv_mask']])
|
|
|
|
orig_img = np.argmax(batch['uv_curvature'][i].numpy(), axis=-1)
|
|
pred_img = np.argmax(pred.numpy(), axis=-1)
|
|
img = np.concatenate((orig_img, np.concatenate(pred_img, axis=1)), axis=1)
|
|
|
|
colors = np.array([cm.color('jet', 1 - i) for i in np.linspace(0, 1, 6+1)])
|
|
color_img = np.zeros((img.shape[0], img.shape[1], 3), np.uint8)
|
|
for i in range(7):
|
|
color_img[img == i, :] = colors[i]
|
|
ex.imshow(color_img).show()
|
|
print()
|
|
print()
|
|
|
|
def sample_2(batch):
|
|
nr_of_samples = 4
|
|
|
|
# pred = model.sample(batch['uv_static'], batch['uv_body_posed'], batch['uv_mask'])
|
|
# pred = model.sample_single(batch['uv_static'][0], batch['uv_body_posed'][0], batch['uv_mask'][0], nr_of_samples=nr_of_samples)
|
|
# pred = model.sample(tf.tile(batch['uv_static'][0][None,:,:,:], [nr_of_samples,1,1,1]), tf.tile(batch['uv_body_posed'][0][None,:,:,:], [nr_of_samples,1,1,1]), tf.tile(batch['uv_mask'][0][None,:,:], [nr_of_samples,1,1]))
|
|
pred,_,_ = model([tf.tile(batch['uv_static'][0][None,:,:,:], [nr_of_samples,1,1,1]), tf.tile(batch['uv_body_posed'][0][None,:,:,:], [nr_of_samples,1,1,1]),tf.tile(batch['uv_curvature'][0][None,:,:,:], [nr_of_samples,1,1,1]), tf.tile(batch['uv_mask'][0][None,:,:], [nr_of_samples,1,1])], training=False)
|
|
# pred = model.call_sample([batch['uv_static'], batch['uv_body_posed'], batch['uv_curvature'], batch['uv_mask']])
|
|
|
|
orig_img = np.argmax(batch['uv_curvature'][0].numpy(), axis=-1)
|
|
pred_img = np.argmax(pred.numpy(), axis=-1)
|
|
img = np.concatenate((orig_img, np.concatenate(pred_img, axis=1)), axis=1)
|
|
|
|
colors = np.array([cm.color('jet', 1 - i) for i in np.linspace(0, 1, 6+1)])
|
|
color_img = np.zeros((img.shape[0], img.shape[1], 3), np.uint8)
|
|
for i in range(7):
|
|
color_img[img == i, :] = colors[i]
|
|
ex.imshow(color_img).show()
|
|
print()
|
|
|
|
|
|
if __name__ == '__main__':
|
|
# init_wandb()
|
|
parser = argparse.ArgumentParser(description='')
|
|
parser.add_argument('--name', type=str)
|
|
args = vars(parser.parse_args())
|
|
|
|
smpl = SMPL()
|
|
skirt_modifier = '_skirt' if config['is_skirt'] else ''
|
|
scce = tf.keras.losses.SparseCategoricalCrossentropy()
|
|
|
|
factory = DatasetFactory()
|
|
|
|
dataset_train = factory.get_dataset(folder=config['train_dir'],
|
|
batch_size=config['batch_size'],
|
|
signature=config['signature_dynamic'],
|
|
resolutions=[config['resolution']],
|
|
in_memory=config['dataset_in_memory'],
|
|
use_one_hot=False,
|
|
use_no_orient=True,
|
|
frames_file=config['frames_file'] if 'frames_file' in config.keys() else None)
|
|
static_dataset_train = StaticDataset(folder=config['train_dir'],
|
|
signature=config['signature_static'],
|
|
resolutions=[config['resolution']],
|
|
in_memory=True)
|
|
|
|
dataset_valid = factory.get_dataset(folder=config['valid_dir'],
|
|
batch_size=config['batch_size'],
|
|
signature=config['signature_dynamic'],
|
|
resolutions=[config['resolution']],
|
|
in_memory=config['dataset_in_memory'],
|
|
use_one_hot=False,
|
|
use_no_orient=True)
|
|
static_dataset_valid = StaticDataset(folder=config['valid_dir'],
|
|
signature=config['signature_static'],
|
|
resolutions=[config['resolution']],
|
|
in_memory=True)
|
|
|
|
model = CVAE(config)
|
|
|
|
sample()
|
|
|
|
# lr_schedule = tf.keras.optimizers.schedules.ExponentialDecay(0.0001, decay_steps=30000, decay_rate=0.1, staircase=False)
|
|
optimizer = tf.optimizers.Adam(learning_rate=config['lr']) #0.0001
|
|
g_balancing_coeff = config['starting_balancing_coeff']
|
|
folder = 'models/ae_logs/'
|
|
valid_iterator = iter(dataset_valid)
|
|
|
|
ckpt = tf.train.Checkpoint(model=model)
|
|
manager = tf.train.CheckpointManager(ckpt, 'checkpoints/iccv/' + config['name'], max_to_keep=20)
|
|
train_accumulator = Accumulator()
|
|
valid_accumulator = Accumulator()
|
|
|
|
if config['pretrained_model'] != '':
|
|
load_model(model, config['pretrained_model'], iteration=config['pretrained_model_iteration'])
|
|
|
|
step = 0
|
|
for batch in dataset_train:
|
|
batch = static_dataset_train.expand_batch(batch_to_dict(batch, config['signature_dynamic']))
|
|
|
|
train_loss = train_step(batch)
|
|
train_accumulator.update(train_loss)
|
|
|
|
if step % 25 == 0:
|
|
data_valid = valid_iterator.get_next()
|
|
data_valid = static_dataset_valid.expand_batch(batch_to_dict(data_valid, config['signature_dynamic']))
|
|
valid_loss = valid_step(data_valid)
|
|
train_accumulator.update(train_loss)
|
|
|
|
if step % 1000 == 0 or (step < 1000 and step % 100 == 0):
|
|
to_log = dict({ **train_loss, **valid_loss, **{ 'current_step': step } })
|
|
if step % 5000 == 0 or (step < 5000 and step % 1000 == 0):
|
|
train_figs = plot_everything(batch, 'train')
|
|
valid_figs = plot_everything(data_valid, 'valid')
|
|
to_log = { **to_log, **train_figs, **valid_figs }
|
|
if step == 0:
|
|
to_log = { **to_log, **{ 'model/summary': wandb.Html(get_summary_html(model)) } }
|
|
|
|
to_log = { **to_log, **{ 'train/balancing_coeff': g_balancing_coeff } }
|
|
|
|
wandb.log(to_log, step=step)
|
|
train_accumulator.reset()
|
|
valid_accumulator.reset()
|
|
|
|
if step % 10000 == 0:
|
|
manager.save(checkpoint_number=step)
|
|
|
|
if step > 40000:
|
|
g_balancing_coeff = g_balancing_coeff + 1 / 20000
|
|
|
|
print(step)
|
|
|
|
step += 1 |