def forward(self, voxel_features, coors, batch_size):
# x = x.contiguous()
coors = coors.int()
# import pdb
# pdb.set_trace()
ret = spconv.SparseConvTensor(voxel_features, coors, self.sparse_shape,
batch_size)
ret = self.downCntx(ret)
down1c, down1b = self.resBlock2(ret)
down2c, down2b = self.resBlock3(down1c)
down3c, down3b = self.resBlock4(down2c)
down4c, down4b = self.resBlock5(down3c)
up4e = self.upBlock0(down4c, down4b)
up3e = self.upBlock1(up4e, down3b)
up2e = self.upBlock2(up3e, down2b)
up1e = self.upBlock3(up2e, down1b)
up0e = self.ReconNet(up1e)
up0e = up0e.replace_feature(
torch.cat((up0e.features, up1e.features), 1))
logits = self.logits(up0e)
y = logits.dense()
return y
def forward(self, features, coors, batch_size, enable_timer: bool = False):
x = spconv.SparseConvTensor(features,
coors,
self.shape,
batch_size,
self.grid,
enable_timer=enable_timer)
return self.net(x)
def forward(self, x: spconv.SparseConvTensor):
inds = x.indices
spatial_shape = [x.batch_size, *x.spatial_shape]
spatial_stride = [0] * len(spatial_shape)
val = 1
for i in range(inds.shape[1] - 1, -1, -1):
spatial_stride[i] = val
val *= spatial_shape[i]
indices_index = inds[:, -1]
for i in range(len(spatial_shape) - 1):
indices_index += spatial_stride[i] * inds[:, i]
_, unique_inds = torch.unique(indices_index)
new_inds = inds[unique_inds]
new_features = x.features[unique_inds]
res = spconv.SparseConvTensor(new_features, new_inds, x.spatial_shape,
x.batch_size, x.grid)
return res
def forward(self, features, coors, batch_size):
x = spconv.SparseConvTensor(features, coors, self.shape, batch_size,
self.grid)
return self.net(x)
def forward(self, features, coors, batch_size):
coors = coors.int()
x = spconv.SparseConvTensor(features, coors, self.shape, batch_size)
return self.todense(self.net(x)) # .dense()