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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371
|
from __future__ import absolute_import, division, print_function
import numpy as np
import torch import torch.nn as nn import torch.nn.functional as F
def disp_to_depth(disp, min_depth, max_depth): """ 将网络的 sigmoid 输出转换为深度预测 该转换公式在论文的“额外考虑”部分给出。 """ min_disp = 1 / max_depth max_disp = 1 / min_depth scaled_disp = min_disp + (max_disp - min_disp) * disp depth = 1 / scaled_disp return scaled_disp, depth
def transformation_from_parameters(axisangle, translation, invert=False): """将网络的 (轴角, 平移) 输出转换为 4x4 矩阵 """ R = rot_from_axisangle(axisangle) t = translation.clone()
if invert: R = R.transpose(1, 2) t *= -1
T = get_translation_matrix(t)
if invert: M = torch.matmul(R, T) else: M = torch.matmul(T, R)
return M
def get_translation_matrix(translation_vector): """将平移向量转换为 4x4 变换矩阵 """ T = torch.zeros(translation_vector.shape[0], 4, 4).to(device=translation_vector.device)
t = translation_vector.contiguous().view(-1, 3, 1)
T[:, 0, 0] = 1 T[:, 1, 1] = 1 T[:, 2, 2] = 1 T[:, 3, 3] = 1 T[:, :3, 3, None] = t
return T
def rot_from_axisangle(vec): """将轴角旋转表示转换为 4x4 变换矩阵 (改编自 https://github.com/Wallacoloo/printipi) 输入 'vec' 必须是 Bx1x3 的形状 """ angle = torch.norm(vec, 2, 2, True) axis = vec / (angle + 1e-7)
ca = torch.cos(angle) sa = torch.sin(angle) C = 1 - ca
x = axis[..., 0].unsqueeze(1) y = axis[..., 1].unsqueeze(1) z = axis[..., 2].unsqueeze(1)
xs = x * sa ys = y * sa zs = z * sa xC = x * C yC = y * C zC = z * C xyC = x * yC yzC = y * zC zxC = z * xC
rot = torch.zeros((vec.shape[0], 4, 4)).to(device=vec.device)
rot[:, 0, 0] = torch.squeeze(x * xC + ca) rot[:, 0, 1] = torch.squeeze(xyC - zs) rot[:, 0, 2] = torch.squeeze(zxC + ys) rot[:, 1, 0] = torch.squeeze(xyC + zs) rot[:, 1, 1] = torch.squeeze(y * yC + ca) rot[:, 1, 2] = torch.squeeze(yzC - xs) rot[:, 2, 0] = torch.squeeze(zxC - ys) rot[:, 2, 1] = torch.squeeze(yzC + xs) rot[:, 2, 2] = torch.squeeze(z * zC + ca) rot[:, 3, 3] = 1
return rot
class ConvBlock(nn.Module): """执行卷积后接 ELU 的层 """ def __init__(self, in_channels, out_channels): super(ConvBlock, self).__init__()
self.conv = Conv3x3(in_channels, out_channels) self.nonlin = nn.ELU(inplace=True) self.bn = nn.BatchNorm2d(out_channels)
def forward(self, x): out = self.conv(x) out = self.bn(out) out = self.nonlin(out) return out
class Conv3x3(nn.Module): """对输入进行填充和卷积的层 """ def __init__(self, in_channels, out_channels, use_refl=True): super(Conv3x3, self).__init__()
if use_refl: self.pad = nn.ReflectionPad2d(1) else: self.pad = nn.ZeroPad2d(1) self.conv = nn.Conv2d(int(in_channels), int(out_channels), 3)
def forward(self, x): out = self.pad(x) out = self.conv(out) return out
class BackprojectDepth(nn.Module): """将深度图像转换为点云的层 """ def __init__(self, batch_size, height, width): super(BackprojectDepth, self).__init__()
self.batch_size = batch_size self.height = height self.width = width
meshgrid = np.meshgrid(range(self.width), range(self.height), indexing='xy') self.id_coords = np.stack(meshgrid, axis=0).astype(np.float32) self.id_coords = nn.Parameter(torch.from_numpy(self.id_coords), requires_grad=False)
self.ones = nn.Parameter(torch.ones(self.batch_size, 1, self.height * self.width), requires_grad=False)
self.pix_coords = torch.unsqueeze(torch.stack( [self.id_coords[0].view(-1), self.id_coords[1].view(-1)], 0), 0) self.pix_coords = self.pix_coords.repeat(batch_size, 1, 1) self.pix_coords = nn.Parameter(torch.cat([self.pix_coords, self.ones], 1), requires_grad=False)
def forward(self, depth, inv_K): cam_points = torch.matmul(inv_K[:, :3, :3], self.pix_coords) cam_points = depth.view(self.batch_size, 1, -1) * cam_points cam_points = torch.cat([cam_points, self.ones], 1)
return cam_points
class Project3D(nn.Module): """将 3D 点投影到具有内参 K 和位置 T 的相机中的层 """ def __init__(self, batch_size, height, width, eps=1e-7): super(Project3D, self).__init__()
self.batch_size = batch_size self.height = height self.width = width self.eps = eps
def forward(self, points, K, T): P = torch.matmul(K, T)[:, :3, :]
cam_points = torch.matmul(P, points)
pix_coords = cam_points[:, :2, :] / (cam_points[:, 2, :].unsqueeze(1) + self.eps) pix_coords = pix_coords.view(self.batch_size, 2, self.height, self.width) pix_coords = pix_coords.permute(0, 2, 3, 1) pix_coords[..., 0] /= self.width - 1 pix_coords[..., 1] /= self.height - 1 pix_coords = (pix_coords - 0.5) * 2 return pix_coords
def upsample(x): """将输入张量上采样 2 倍 """ return F.interpolate(x, scale_factor=2, mode="nearest")
def get_smooth_loss(disp, img): """计算视差图像的平滑损失 彩色图像用于边缘感知平滑 """ grad_disp_x = torch.abs(disp[:, :, :, :-1] - disp[:, :, :, 1:]) grad_disp_y = torch.abs(disp[:, :, :-1, :] - disp[:, :, 1:, :])
grad_img_x = torch.mean(torch.abs(img[:, :, :, :-1] - img[:, :, :, 1:]), 1, keepdim=True) grad_img_y = torch.mean(torch.abs(img[:, :, :-1, :] - img[:, :, 1:, :]), 1, keepdim=True)
grad_disp_x *= torch.exp(-grad_img_x) grad_disp_y *= torch.exp(-grad_img_y)
return grad_disp_x.mean() + grad_disp_y.mean()
class SSIM(nn.Module): """计算一对图像之间 SSIM 损失的层 """ def __init__(self): super(SSIM, self).__init__() self.mu_x_pool = nn.AvgPool2d(3, 1) self.mu_y_pool = nn.AvgPool2d(3, 1) self.sig_x_pool = nn.AvgPool2d(3, 1) self.sig_y_pool = nn.AvgPool2d(3, 1) self.sig_xy_pool = nn.AvgPool2d(3, 1)
self.refl = nn.ReflectionPad2d(1)
self.C1 = 0.01 ** 2 self.C2 = 0.03 ** 2
def forward(self, x, y): x = self.refl(x) y = self.refl(y)
mu_x = self.mu_x_pool(x) mu_y = self.mu_y_pool(y)
sigma_x = self.sig_x_pool(x ** 2) - mu_x ** 2 sigma_y = self.sig_y_pool(y ** 2) - mu_y ** 2 sigma_xy = self.sig_xy_pool(x * y) - mu_x * mu_y
SSIM_n = (2 * mu_x * mu_y + self.C1) * (2 * sigma_xy + self.C2) SSIM_d = (mu_x ** 2 + mu_y ** 2 + self.C1) * (sigma_x + sigma_y + self.C2)
return torch.clamp((1 - SSIM_n / SSIM_d) / 2, 0, 1)
def compute_depth_errors(gt, pred): """计算预测深度和真实深度之间的误差指标 """ thresh = torch.max((gt / pred), (pred / gt)) a1 = (thresh < 1.25 ).float().mean() a2 = (thresh < 1.25 ** 2).float().mean() a3 = (thresh < 1.25 ** 3).float().mean()
rmse = (gt - pred) ** 2 rmse = torch.sqrt(rmse.mean())
rmse_log = (torch.log(gt) - torch.log(pred)) ** 2 rmse_log = torch.sqrt(rmse_log.mean())
abs_rel = torch.mean(torch.abs(gt - pred) / gt)
sq_rel = torch.mean((gt - pred) ** 2 / gt)
return abs_rel, sq_rel, rmse, rmse_log, a
|