def build_losses(self, losses_cfg):
self.cls_loss_func = loss_utils.SigmoidFocalClassificationLoss_v1(
alpha=0.25, gamma=2.0)
self.reg_loss_func = F.smooth_l1_loss if losses_cfg.get(
'LOSS_REG', None) == 'smooth-l1' else F.l1_loss
print("done")
def __init__(self, inchannel):
super().__init__()
norm_fun = partial(nn.BatchNorm1d, eps=1e-3, momentum=0.01)
block = self.conv_block
self.conv1 = nn.Sequential(
nn.Conv1d(inchannel, 32, 3, bias=False, stride=1, padding=1),
norm_fun(32), nn.ReLU())
self.conv2 = nn.Sequential(
nn.Conv1d(32, 32, 3, bias=False, stride=1, padding=1),
norm_fun(32), nn.ReLU())
self.conv3 = nn.Sequential(
nn.Conv1d(32, 32, 3, bias=False, stride=1, padding=1),
norm_fun(32), nn.ReLU())
self.deconv1 = nn.Sequential(
nn.ConvTranspose1d(32, 64, 3, stride=1, bias=False, padding=1),
norm_fun(64), nn.ReLU())
self.conv4 = nn.Sequential(
nn.Conv1d(32, 32, 3, bias=False, stride=1, padding=1),
norm_fun(32), nn.ReLU())
self.conv5 = nn.Sequential(
nn.Conv1d(32, 32, 3, bias=False, stride=1, padding=1),
norm_fun(32), nn.ReLU())
self.conv6 = nn.Sequential(
nn.Conv1d(32, 32, 3, bias=False, stride=1, padding=1),
norm_fun(32), nn.ReLU())
self.deconv2 = nn.Sequential(
nn.ConvTranspose1d(32, 64, 3, stride=1, bias=False, padding=1),
norm_fun(64), nn.ReLU())
self.cls_layer = nn.Conv1d(128, 2, 1, bias=True)
self.seg_loss_func = loss_utils.SigmoidFocalClassificationLoss_v1(
alpha=0.25, gamma=2.0)
self.ret = {}
self.init_weights()
def __init__(self, in_channels):
super().__init__()
# self.print_info = cfg.print_info
norm_fn = partial(nn.BatchNorm1d, eps=1e-3, momentum=0.01)
self.conv_input = spconv.SparseSequential(
spconv.SubMConv3d(in_channels,
16,
3,
padding=1,
bias=False,
indice_key='subm1'),
norm_fn(16),
nn.ReLU(),
)
block = self.post_act_block
self.conv1 = spconv.SparseSequential(
block(16, 16, 3, norm_fn=norm_fn, padding=1, indice_key='subm1'), )
self.conv2 = spconv.SparseSequential(
# [1600, 1408, 41] <- [800, 704, 21]
block(16,
32,
3,
norm_fn=norm_fn,
stride=2,
padding=1,
indice_key='spconv2',
conv_type='spconv'),
block(32, 32, 3, norm_fn=norm_fn, padding=1, indice_key='subm2'),
block(32, 32, 3, norm_fn=norm_fn, padding=1, indice_key='subm2'),
)
self.conv3 = spconv.SparseSequential(
# [800, 704, 21] <- [400, 352, 11]
block(32,
64,
3,
norm_fn=norm_fn,
stride=2,
padding=1,
indice_key='spconv3',
conv_type='spconv'),
block(64, 64, 3, norm_fn=norm_fn, padding=1, indice_key='subm3'),
block(64, 64, 3, norm_fn=norm_fn, padding=1, indice_key='subm3'),
)
self.conv4 = spconv.SparseSequential(
# [400, 352, 11] <- [200, 176, 5]
block(64,
64,
3,
norm_fn=norm_fn,
stride=2,
padding=(0, 1, 1),
indice_key='spconv4',
conv_type='spconv'),
block(64, 64, 3, norm_fn=norm_fn, padding=1, indice_key='subm4'),
block(64, 64, 3, norm_fn=norm_fn, padding=1, indice_key='subm4'),
)
last_pad = 0 if cfg.DATA_CONFIG.VOXEL_GENERATOR.VOXEL_SIZE[-1] in [
0.1, 0.2
] else (1, 0, 0)
self.conv_out = spconv.SparseSequential(
# [200, 150, 5] -> [200, 150, 2]
spconv.SparseConv3d(64,
128, (3, 1, 1),
stride=(2, 1, 1),
padding=last_pad,
bias=False,
indice_key='spconv_down2'),
norm_fn(128),
nn.ReLU(),
)
self.seg_net = SegNet(16)
self.seg_loss_func = loss_utils.SigmoidFocalClassificationLoss_v1(
alpha=0.25, gamma=2.0)
self.num_point_features = 128
self.ret = {}
def build_losses(self,losses_cfg):
# loss function definition
self.cls_loss_layer = loss_utils.SigmoidFocalClassificationLoss_v1(alpha=0.25, gamma=2.0)
self.reg_loss_layer = loss_utils.WeightedSmoothL1Loss_v1(code_weights=losses_cfg['code_loss_weight'])
self.dir_loss_layer = loss_utils.WeightedCrossEntropyLoss_v1()