Spaces:
Runtime error
Runtime error
File size: 11,511 Bytes
c3d0293 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 |
from torch import nn
import torch
import numpy as np
from SMPLX.visualize_joint2smpl.simplify_loc2rot import joints2smpl
from motion.hybrik_loc2rot import HybrIKJointsToRotmat
from SMPLX import smplx
from SMPLX.read_from_npy import npy2info, info2dict
from SMPLX.rotation_conversions import *
from motion.dataset.recover_smr import *
from motion.dataset.recover_joints import recover_from_ric as recover_joints
from motion.dataset.paramUtil import t2m_kinematic_chain
from motion.plot3d import plot_3d_motion
import os
import subprocess
import platform
from PIL import Image
from motion.sample import Predictor as mdm_predictor
from TADA.anime import Animation
class Visualize(nn.Module):
def __init__(self, **kargs):
super(Visualize, self).__init__()
self.mode = kargs.get("mode", "cadm")
if self.mode in ["mdm", "cadm", "cadm-augment"]:
self.predictor = mdm_predictor(**kargs)
self.rep = self.predictor.rep
self.smpl_path = kargs.get("smpl_path")
self.device = kargs.get("device", "cpu")
self.rotate = kargs.get("rotate", 0)
self.pose_generator = HybrIKJointsToRotmat()
self.path = kargs["path"]
self.tada_base = kargs.get("tada_base", None)
self.tada_role = kargs.get("tada_role", None)
if self.tada_base is not None and self.tada_role is not None:
self.anime = Animation(self.tada_role, self.tada_base, self.device)
self.face = None
else:
self.face = np.load(os.path.join(self.path["dataset_dir"], "smplh.faces"))
self.anime = None
def fit2smpl(self, motion, mode="fast"):
print(">>>>>>>>>>>>>>> fit joints to smpl >>>>>>>>>>>>>>>>>>>>")
if mode == "slow":
frames = motion.shape[0]
j2s = joints2smpl(num_frames=frames, device=self.device, model_path=self.smpl_path, json_dict=self.path)
motion_tensor, translation = j2s.joint2smpl(motion)
else:
translation = motion[:, 0:1, :] - motion[0, 0:1, :]
motion = self.pose_generator(motion)
motion = torch.from_numpy(motion)
hand = torch.eye(3).unsqueeze(0).unsqueeze(0).repeat(motion.shape[0], 2, 1, 1)
motion = torch.cat([motion, hand], dim=1)
motion_tensor = matrix_to_axis_angle(motion)
motion_tensor = motion_tensor.numpy()
return motion_tensor, translation
def predict(self, sentence, path, render_mode="pyrender", joint_path=None, smpl_path=None):
if self.mode == "pose":
motion_tensor = np.load(path)
if render_mode == "joints":
_, joints = self.get_mesh(motion_tensor)
motion_tensor = joints
elif self.mode == "joints":
joints = np.load(path)
if render_mode == "joints":
motion_tensor = joints
else:
motion_tensor, translation = self.fit2smpl(joints, render_mode.split("_")[-1])
motion_tensor = np.concatenate([motion_tensor, translation], axis=1)
motion_tensor = motion_tensor.reshape(motion_tensor.shape[0], -1)
elif self.mode in ["mdm", "cadm", "cadm-augment"]:
motion_tensor = self.predictor.predict(sentence, 1, path)
if self.rep == "t2m":
motion_tensor = motion_tensor[0].detach().cpu().numpy() #### [nframes, 263]
if joint_path is not None:
np.save(joint_path, motion_tensor)
if render_mode == "joints":
motion_tensor = motion_tensor
else:
motion_tensor, translation = self.fit2smpl(motion_tensor, render_mode.split("_")[-1])
motion_tensor = np.concatenate([motion_tensor, translation], axis=1)
motion_tensor = motion_tensor.reshape(motion_tensor.shape[0], -1)
if smpl_path is not None:
np.save(smpl_path, motion_tensor)
elif self.rep == "smr":
motion_tensor = motion_tensor[0][0].detach().cpu().numpy()
joints = recover_from_ric(motion_tensor, 22)
if joint_path is not None:
np.save(joint_path, joints)
if render_mode == "joints":
motion_tensor = joints
else:
pose = recover_pose_from_smr(motion_tensor, 22)
pose = pose.reshape(pose.shape[0], -1, 3)
motion_tensor, translation = self.fit2smpl(joints, render_mode.split("_")[-1])
motion_tensor = np.concatenate([motion_tensor, translation], axis=1)
motion_tensor = motion_tensor.reshape(motion_tensor.shape[0], -1, 3)
replace = [12, 15, 20, 21]
motion_tensor[:, replace, :] = pose[:, replace, :]
motion_tensor = motion_tensor.reshape(motion_tensor.shape[0], -1)
if smpl_path is not None:
np.save(smpl_path, motion_tensor)
return motion_tensor.astype(np.float32)
def joints_process(self, joints, text, width=1024, height=1024):
os.makedirs("temp", exist_ok=True)
plot_3d_motion(t2m_kinematic_chain, joints, text, figsize=(width/100, height/100))
files = os.listdir("temp")
files = sorted(files)
pics = []
for i in range(len(files)):
pic = Image.open(os.path.join("temp", files[i]))
pic = np.asarray(pic)
pics.append(pic.copy())
cmd = "rm -r temp"
subprocess.call(cmd, shell=platform.system() != 'Windows')
pics = np.stack(pics, axis=0)
return pics
def pyrender_process(self, vertices, height=1024, weight=1024):
import trimesh
from trimesh import Trimesh
import pyrender
from pyrender.constants import RenderFlags
import os
os.environ['PYOPENGL_PLATFORM'] = "egl"
from shapely import geometry
from tqdm import tqdm
faces = self.face
vertices = vertices.astype(np.float32)
MINS = np.min(np.min(vertices, axis=0), axis=0)
MAXS = np.max(np.max(vertices, axis=0), axis=0)
#################### position initial at zero point
vertices[:, :, 0] -= (MAXS + MINS)[0] / 2
vertices[:, :, 2] -= (MAXS + MINS)[2] / 2
MINS = np.min(np.min(vertices, axis=0), axis=0)
MAXS = np.max(np.max(vertices, axis=0), axis=0)
pics = []
############### ground initial ###########
minx = MINS[0] - 0.5
maxx = MAXS[0] + 0.5
minz = MINS[2] - 0.5
maxz = MAXS[2] + 0.5
polygon = geometry.Polygon([[minx, minz], [minx, maxz], [maxx, maxz], [maxx, minz]])
polygon_mesh = trimesh.creation.extrude_polygon(polygon, 1e-5)
polygon_mesh.visual.face_colors = [0, 0, 0, 0.21]
polygon_render = pyrender.Mesh.from_trimesh(polygon_mesh, smooth=False)
r = pyrender.OffscreenRenderer(weight, height)
for i in tqdm(range(vertices.shape[0])):
end_color = np.array([30, 128, 255]) / 255.0
bg_color = [1, 1, 1, 0.8]
scene = pyrender.Scene(bg_color=bg_color, ambient_light=(0.4, 0.4, 0.4))
if self.anime is None:
mesh = Trimesh(vertices=vertices[i, :, :].tolist(), faces=faces)
base_color = end_color.tolist()
material = pyrender.MetallicRoughnessMaterial(
metallicFactor=0.7, roughnessFactor=0.7,
alphaMode='OPAQUE',
baseColorFactor=base_color
)
mesh = pyrender.Mesh.from_trimesh(mesh, material=material)
else:
mesh = Trimesh(vertices=vertices[i, :, :].tolist(), faces=faces, visual=self.anime.trimesh_visual, process=False)
mesh = pyrender.Mesh.from_trimesh(mesh, smooth=True, material=None)
scene.add(mesh)
########################### ground ##################
c = np.pi / 2
scene.add(polygon_render, pose=np.array([[ 1, 0, 0, 0],
[ 0, np.cos(c), -np.sin(c), MINS[1]],
[ 0, np.sin(c), np.cos(c), 0],
[ 0, 0, 0, 1]]))
################ light ############
light = pyrender.DirectionalLight(color=[1,1,1], intensity=300)
light_pose = np.eye(4)
light_pose[:3, 3] = [0, -1, 1]
scene.add(light, pose=light_pose.copy())
light_pose[:3, 3] = [0, 1, 1]
scene.add(light, pose=light_pose.copy())
light_pose[:3, 3] = [1, 1, 2]
scene.add(light, pose=light_pose.copy())
################ camera ##############
camera = pyrender.PerspectiveCamera(yfov=(np.pi / 3.0))
c = -np.pi / 6
scene.add(camera, pose=[[ 1, 0, 0, (minx+maxx)],
[ 0, np.cos(c), -np.sin(c), 2.5],
[ 0, np.sin(c), np.cos(c), max(4, minz+(1.5-MINS[1])*2, (maxx-minx))],
[ 0, 0, 0, 1]
])
pic, _ = r.render(scene, flags=RenderFlags.RGBA)
pics.append(pic)
pics = np.stack(pics, axis=0)
return pics
@torch.no_grad()
def get_mesh(self, motions):
if self.anime is not None:
vertices, faces = self.anime.forward_mdm(motions)
joints = vertices
self.face = faces
else:
motions, trans, gender, betas = npy2info(motions, 10)
betas = None
gender = "neutral"
if motions.shape[1] == 72:
mode = "smpl"
elif motions.shape[1] == 156:
mode = "smplh"
elif motions.shape[1] == 165:
motions = np.concatenate([motions[:, :66], motions[:, 75::]], axis=1)
mode = "smplh"
if self.rotate != 0:
motions = motions.reshape(motions.shape[0], -1, 3)
motions = torch.from_numpy(motions).float()
first_frame_root_pose_matrix = axis_angle_to_matrix(motions[0][0])
all_root_poses_matrix = axis_angle_to_matrix(motions[:, 0, :])
aligned_root_poses_matrix = torch.matmul(torch.transpose(first_frame_root_pose_matrix, 0, 1),
all_root_poses_matrix)
motions[:, 0, :] = matrix_to_axis_angle(aligned_root_poses_matrix)
motions = motions.reshape(motions.shape[0], -1)
motions = motions.numpy()
print("Visualize Mode -> ", mode)
model = smplx.create(self.smpl_path, model_type=mode,
gender=gender, use_face_contour=True,
num_betas=10,
num_expression_coeffs=10,
ext="npz", use_pca=False, batch_size=motions.shape[0])
model = model.eval().to(self.device)
inputs = info2dict(motions, trans, betas, mode, self.device)
output = model(**inputs)
vertices = output.vertices.cpu().numpy()
joints = output.joints.cpu().numpy()
return vertices, joints |