confused about Dilated convolution replace maxpooling

[chasingw]

hello, I'm confused about this change in class PillarFeatureNet

nutonomy_pointpillars/second/pytorch/models/pointpillars.py

Lines 52 to 73 in 68d3181

	self.t_conv = nn.ConvTranspose2d(100, 1, (1,8), stride=(1,7))
	self.conv3 = nn.Conv2d(64, 64, kernel_size=(1, 34), stride=(1, 1), dilation=(1,3))
	def forward(self, input):
	x = self.conv1(input)
	x = self.norm(x)
	x = F.relu(x)
	x = self.conv3(x)
	return x
	# x = self.linear(input)
	# x = self.norm(x.permute(0, 2, 1).contiguous()).permute(0, 2, 1).contiguous()
	# x = F.relu(x)
	#
	# x_max = torch.max(x, dim=1, keepdim=True)[0]
	#
	# if self.last_vfe:
	# return x_max
	# else:
	# x_repeat = x_max.repeat(1, inputs.shape[1], 1)
	# x_concatenated = torch.cat([x, x_repeat], dim=2)
	# return x_concatenated

use conv replace maxpooling op
although the shape are same, can this change equivalent? do you know some reference material about this

[ldtho]

same question!