File size: 14,979 Bytes
2c68f56 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 | #!/usr/bin/env python
# Copyright (c) Xuangeng Chu (xg.chu@outlook.com)
import os
import torch
import numpy as np
import torch.nn as nn
from pytorch3d.io import load_obj
from pytorch3d.structures import (
Meshes, Pointclouds
)
from pytorch3d.renderer import (
PerspectiveCameras, OrthographicCameras,
look_at_view_transform, RasterizationSettings,
FoVPerspectiveCameras, PointsRasterizationSettings,
PointsRenderer, PointsRasterizer, AlphaCompositor,
PointLights, AmbientLights, TexturesVertex, TexturesUV, BlendParams,
SoftPhongShader, MeshRasterizer, MeshRenderer, SoftSilhouetteShader, HardFlatShader
)
from pytorch3d.transforms import matrix_to_rotation_6d, rotation_6d_to_matrix
from .util import weak_cam2persp_cam
class BaseMeshRenderer(nn.Module):
def __init__(self, assets_dir, image_size=512, device='cuda', skin_color=[252, 224, 203], bg_color=[0, 0, 0], focal_length=12):
super(BaseMeshRenderer, self).__init__()
self.device = device
self.image_size = image_size
self.assets_dir = assets_dir
self.skin_color = np.array(skin_color)
self.bg_color = bg_color
self.focal_length = focal_length
self.raster_settings = RasterizationSettings(image_size=image_size, blur_radius=0.0, faces_per_pixel=1)
self.lights = PointLights(device=device, location=[[0.0, 0.0, 3.0]])
self.manual_lights = PointLights(
device=self.device,
location=((0.0, 0.0, 5.0), ),
ambient_color=((0.5, 0.5, 0.5), ),
diffuse_color=((0.5, 0.5, 0.5), ),
specular_color=((0.01, 0.01, 0.01), )
)
self.blend = BlendParams(background_color=np.array(bg_color)/225.)
def _build_cameras(self, transform_matrix, focal_length):
batch_size = transform_matrix.shape[0]
screen_size = torch.tensor(
[self.image_size, self.image_size], device=self.device
).float()[None].repeat(batch_size, 1)
cameras_kwargs = {
'principal_point': torch.zeros(batch_size, 2, device=self.device).float(), 'focal_length': focal_length,
'image_size': screen_size, 'device': self.device,
}
cameras = PerspectiveCameras(**cameras_kwargs, R=transform_matrix[:, :3, :3], T=transform_matrix[:, :3, 3])
return cameras
def _build_orth_cameras(self):
R, T = look_at_view_transform(dist=10)
return OrthographicCameras(device=self.device, focal_length=1, R=R, T=T,
image_size=torch.tensor([self.image_size, self.image_size], device=self.device).float())
def forward(self, vertices, faces=None, landmarks={}, cameras=None, transform_matrix=None, focal_length=None, is_weak_cam=False, ret_image=True):
B, V = vertices.shape[:2]
focal_length = self.focal_length if focal_length is None else focal_length
if isinstance(cameras, torch.Tensor):
cameras = cameras.clone()
elif is_weak_cam:
cameras = self._build_orth_cameras()
elif cameras is None:
cameras = self._build_cameras(transform_matrix, focal_length)
t_faces = faces[None].repeat(B, 1, 1)
ret_vertices = cameras.transform_points_screen(vertices)
ret_landmarks = {k: cameras.transform_points_screen(v) for k,v in landmarks.items()}
images = None
if ret_image:
# Initialize each vertex to be white in color.
verts_rgb = torch.from_numpy(self.skin_color/255).float().to(self.device)[None, None, :].repeat(B, V, 1)
textures = TexturesVertex(verts_features=verts_rgb)
mesh = Meshes(
verts=vertices.to(self.device),
faces=t_faces.to(self.device),
textures=textures
)
renderer = MeshRenderer(
rasterizer=MeshRasterizer(cameras=cameras, raster_settings=self.raster_settings),
shader=SoftPhongShader(cameras=cameras, lights=self.lights, device=self.device, blend_params=self.blend)
)
render_results = renderer(mesh).permute(0, 3, 1, 2)
images = render_results[:, :3]
alpha_images = render_results[:, 3:]
images[alpha_images.expand(-1, 3, -1, -1)<0.5] = 0.0
images = images * 255
return ret_vertices, ret_landmarks, images
def render_mesh(self, vertices,cameras, faces=None,lights=None):
# if not hasattr(self, 'orth_cameras'):
# R, T = look_at_view_transform(dist=10)
# self.orth_cameras = OrthographicCameras(device=self.device, focal_length=1, R=R, T=T)
if faces is None:
faces = self.faces.squeeze(0)
self.lights=lights
if lights is None:
self.lights = PointLights(device=self.device, location=[[0.0, 1.0, 10.0]])
B, V = vertices.shape[:2]
t_faces = faces[None].repeat(B, 1, 1)
images = None
# Initialize each vertex to be white in color.
verts_rgb = torch.from_numpy(self.skin_color/255).float().to(self.device)[None, None, :].repeat(B, V, 1)
textures = TexturesVertex(verts_features=verts_rgb)
mesh = Meshes(
verts=vertices.to(self.device),
faces=t_faces.to(self.device),
textures=textures
)
# renderer = MeshRenderer(
# rasterizer=MeshRasterizer(cameras=self.orth_cameras, raster_settings=self.raster_settings),
# shader=SoftPhongShader(cameras=self.orth_cameras, lights=self.lights, device=self.device, blend_params=self.blend)
# )
renderer = MeshRenderer(
rasterizer=MeshRasterizer(cameras=cameras, raster_settings=self.raster_settings),
shader=SoftPhongShader(cameras=cameras, lights=self.lights, device=self.device, blend_params=self.blend)
)
render_results = renderer(mesh).permute(0, 3, 1, 2)
images = render_results[:, :3]
alpha_images = render_results[:, 3:]
images[alpha_images.expand(-1, 3, -1, -1)<0.5] = 0.0
images = images * 255
return images
def render_alpha(self, vertices,cameras, faces=None,lights=None):
if faces is None:
faces = self.faces.squeeze(0)
self.lights=lights
if lights is None:
self.lights = PointLights(device=self.device, location=[[0.0, 1.0, 10.0]])
B, V = vertices.shape[:2]
t_faces = faces[None].repeat(B, 1, 1)
images = None
# Initialize each vertex to be white in color.
verts_rgb = torch.from_numpy(self.skin_color/255).float().to(self.device)[None, None, :].repeat(B, V, 1)
textures = TexturesVertex(verts_features=verts_rgb)
mesh = Meshes(
verts=vertices.to(self.device),
faces=t_faces.to(self.device),
textures=textures
)
renderer = MeshRenderer(
rasterizer=MeshRasterizer(cameras=cameras, raster_settings=self.raster_settings),
shader=SoftPhongShader(cameras=cameras, lights=self.lights, device=self.device, blend_params=self.blend)
)
render_results = renderer(mesh).permute(0, 3, 1, 2)
alpha_images = render_results[:, 3:]
return alpha_images
class PointRenderer(nn.Module):
def __init__(self, image_size=256, device='cpu'):
super(PointRenderer, self).__init__()
self.device = device
R, T = look_at_view_transform(4, 30, 30) # d, e, a
self.cameras = FoVPerspectiveCameras(device=device, R=R, T=T, znear=0.01, zfar=1.0)
raster_settings = PointsRasterizationSettings(
image_size=image_size, radius=0.005, points_per_pixel=10
)
rasterizer = PointsRasterizer(cameras=self.cameras, raster_settings=raster_settings)
self.renderer = PointsRenderer(rasterizer=rasterizer, compositor=AlphaCompositor())
def forward(self, points, D=3, E=15, A=30, coords=True, ex_points=None):
if D !=8 or E != 30 or A != 30:
R, T = look_at_view_transform(D, E, A) # d, e, a
self.cameras = FoVPerspectiveCameras(device=self.device, R=R, T=T, znear=0.01, zfar=1.0)
verts = torch.Tensor(points).to(self.device)
verts = verts[:, torch.randperm(verts.shape[1])[:10000]]
if ex_points is not None:
verts = torch.cat([verts, ex_points.expand(verts.shape[0], -1, -1)], dim=1)
if coords:
coords_size = verts.shape[1]//10
cod = verts.new_zeros(coords_size*3, 3)
li = torch.linspace(0, 1.0, steps=coords_size, device=cod.device)
cod[:coords_size, 0], cod[coords_size:coords_size*2, 1], cod[coords_size*2:, 2] = li, li, li
verts = torch.cat(
[verts, cod.unsqueeze(0).expand(verts.shape[0], -1, -1)], dim=1
)
rgb = torch.Tensor(torch.rand_like(verts)).to(self.device)
point_cloud = Pointclouds(points=verts, features=rgb)
images = self.renderer(point_cloud, cameras=self.cameras,).permute(0, 3, 1, 2)
return images*255
class TextureRenderer(nn.Module):
def __init__(self, obj_filename=None, tuv=None, flame_mask=None, device='cpu'):
super(TextureRenderer, self).__init__()
self.device = device
# objects
if obj_filename is not None:
_, faces, aux = load_obj(obj_filename, load_textures=False)
self.uvverts = aux.verts_uvs[None, ...].to(self.device) # (N, V, 2)
self.uvfaces = faces.textures_idx[None, ...].to(self.device) # (N, F, 3)
self.faces = faces.verts_idx[None, ...].to(self.device) # (N, F, 3)
elif tuv is not None:
import numpy as np
self.uvverts = tuv['verts_uvs'][None, ...].to(self.device) # (N, V, 2)
self.uvfaces = tuv['textures_idx'][None, ...].to(self.device) # (N, F, 3)
self.faces = tuv['verts_idx'][None, ...].to(self.device) # (N, F, 3)
else:
raise NotImplementedError('Must have faces and uvs.')
# setting
self.lights = AmbientLights(device=self.device)
# flame mask
if flame_mask is not None:
reduced_faces = []
for f in self.faces[0]:
valid = 0
for v in f:
if v.item() in flame_mask:
valid += 1
reduced_faces.append(True if valid == 3 else False)
reduced_faces = torch.tensor(reduced_faces).to(self.device)
self.flame_mask = reduced_faces
## lighting
pi = np.pi
sh_const = torch.tensor(
[
1 / np.sqrt(4 * pi),
((2 * pi) / 3) * (np.sqrt(3 / (4 * pi))),
((2 * pi) / 3) * (np.sqrt(3 / (4 * pi))),
((2 * pi) / 3) * (np.sqrt(3 / (4 * pi))),
(pi / 4) * (3) * (np.sqrt(5 / (12 * pi))),
(pi / 4) * (3) * (np.sqrt(5 / (12 * pi))),
(pi / 4) * (3) * (np.sqrt(5 / (12 * pi))),
(pi / 4) * (3 / 2) * (np.sqrt(5 / (12 * pi))),
(pi / 4) * (1 / 2) * (np.sqrt(5 / (4 * pi))),
],
dtype=torch.float32,
)
self.constant_factor = sh_const.to(self.device)
def add_SHlight(self, normal_images, sh_coeff):
# sh_coeff: [bz, 9, 3]
N = normal_images
sh = torch.stack([
N[:, 0] * 0. + 1., N[:, 0], N[:, 1],
N[:, 2], N[:, 0] * N[:, 1], N[:, 0] * N[:, 2],
N[:, 1] * N[:, 2], N[:, 0] ** 2 - N[:, 1] ** 2, 3 * (N[:, 2] ** 2) - 1
], 1) # [bz, 9, h, w]
sh = sh * self.constant_factor[None, :, None, None]
shading = torch.sum(sh_coeff[:, :, :, None, None] * sh[:, :, None, :, :], 1) # [bz, 9, 3, h, w]
return shading
def _build_cameras(self, transform_matrix, focal_length, principal_point, image_size):
batch_size = transform_matrix.shape[0]
screen_size = torch.tensor(
[image_size, image_size], device=self.device
).float()[None].repeat(batch_size, 1)
cameras_kwargs = {
'principal_point': principal_point.repeat(batch_size, 1), 'focal_length': focal_length,
'image_size': screen_size, 'device': self.device,
}
cameras = PerspectiveCameras(**cameras_kwargs, R=transform_matrix[:, :3, :3], T=transform_matrix[:, :3, 3])
return cameras
def forward(
self, vertices_world, texture_images, lights=None, image_size=512,
cameras=None, transform_matrix=None, focal_length=None, principal_point=None
):
if cameras is None:
cameras = self._build_cameras(transform_matrix, focal_length, principal_point, image_size)
batch_size = vertices_world.shape[0]
faces = self.faces.expand(batch_size, -1, -1)
textures_uv = TexturesUV(
maps=texture_images.expand(batch_size, -1, -1, -1).permute(0, 2, 3, 1),
faces_uvs=self.uvfaces.expand(batch_size, -1, -1),
verts_uvs=self.uvverts.expand(batch_size, -1, -1)
)
meshes_world = Meshes(verts=vertices_world, faces=faces, textures=textures_uv)
# phong renderer
raster_settings = RasterizationSettings(
image_size=image_size, blur_radius=0.0, faces_per_pixel=1,
perspective_correct=True, cull_backfaces=True
)
phong_renderer = MeshRenderer(
rasterizer=MeshRasterizer(cameras=cameras, raster_settings=raster_settings),
shader=SoftPhongShader(device=self.device, cameras=cameras, lights=self.lights)
)
image_ref = phong_renderer(meshes_world=meshes_world)
images = image_ref[..., :3].permute(0, 3, 1, 2)
masks_all = image_ref[..., 3:].permute(0, 3, 1, 2) > 0.0
if lights is not None:
images = self.add_SHlight(images, lights)
images[~masks_all.expand(-1, 3, -1, -1)] = 0.0
# silhouette renderer
with torch.no_grad():
if hasattr(self, 'flame_mask'):
textures_verts = TexturesVertex(verts_features=vertices_world.new_ones(vertices_world.shape))
meshes_masked = Meshes(
verts=vertices_world, faces=faces[:, self.flame_mask], textures=textures_verts
)
silhouette_renderer = MeshRenderer(
rasterizer=MeshRasterizer(cameras=cameras, raster_settings=raster_settings),
shader=SoftSilhouetteShader()
)
masks_face = silhouette_renderer(meshes_world=meshes_masked)
masks_face = masks_face[..., 3:].permute(0, 3, 1, 2) > 0.0
else:
masks_face = None
return images, masks_all, masks_face
|