diff --git a/pcdet/models/backbones_3d/spconv_backbone_voxelnext2d.py b/pcdet/models/backbones_3d/spconv_backbone_voxelnext2d.py new file mode 100644 index 0000000..bdfdf99 --- /dev/null +++ b/pcdet/models/backbones_3d/spconv_backbone_voxelnext2d.py @@ -0,0 +1,219 @@ +from functools import partial +import torch +import torch.nn as nn + +from ...utils.spconv_utils import replace_feature, spconv + + +def post_act_block(in_channels, out_channels, kernel_size, indice_key=None, stride=1, padding=0, + conv_type='subm', norm_fn=None): + + if conv_type == 'subm': + conv = spconv.SubMConv2d(in_channels, out_channels, kernel_size, bias=False, indice_key=indice_key) + elif conv_type == 'spconv': + conv = spconv.SparseConv2d(in_channels, out_channels, kernel_size, stride=stride, padding=padding, + bias=False, indice_key=indice_key) + elif conv_type == 'inverseconv': + conv = spconv.SparseInverseConv2d(in_channels, out_channels, kernel_size, indice_key=indice_key, bias=False) + else: + raise NotImplementedError + + m = spconv.SparseSequential( + conv, + norm_fn(out_channels), + nn.ReLU(), + ) + + return m + + +class SparseBasicBlock(spconv.SparseModule): + expansion = 1 + + def __init__(self, inplanes, planes, stride=1, norm_fn=None, downsample=None, indice_key=None): + super(SparseBasicBlock, self).__init__() + + assert norm_fn is not None + bias = norm_fn is not None + self.conv1 = spconv.SubMConv2d( + inplanes, planes, kernel_size=3, stride=stride, padding=1, bias=bias, indice_key=indice_key + ) + self.bn1 = norm_fn(planes) + self.relu = nn.ReLU() + self.conv2 = spconv.SubMConv2d( + planes, planes, kernel_size=3, stride=stride, padding=1, bias=bias, indice_key=indice_key + ) + self.bn2 = norm_fn(planes) + self.downsample = downsample + self.stride = stride + + def forward(self, x): + identity = x + + out = self.conv1(x) + out = replace_feature(out, self.bn1(out.features)) + out = replace_feature(out, self.relu(out.features)) + + out = self.conv2(out) + out = replace_feature(out, self.bn2(out.features)) + + if self.downsample is not None: + identity = self.downsample(x) + + out = replace_feature(out, out.features + identity.features) + out = replace_feature(out, self.relu(out.features)) + + return out + + +class VoxelResBackBone8xVoxelNeXt2D(nn.Module): + def __init__(self, model_cfg, input_channels, grid_size, **kwargs): + super().__init__() + self.model_cfg = model_cfg + norm_fn = partial(nn.BatchNorm1d, eps=1e-3, momentum=0.01) + self.sparse_shape = grid_size[[1, 0]] + + block = post_act_block + + spconv_kernel_sizes = model_cfg.get('SPCONV_KERNEL_SIZES', [3, 3, 3, 3]) + + self.conv1 = spconv.SparseSequential( + SparseBasicBlock(32, 32, norm_fn=norm_fn, indice_key='res1'), + SparseBasicBlock(32, 32, norm_fn=norm_fn, indice_key='res1'), + SparseBasicBlock(32, 32, norm_fn=norm_fn, indice_key='res1'), + ) + + self.conv2 = spconv.SparseSequential( + # [1600, 1408] <- [800, 704] + block(32, 64, spconv_kernel_sizes[0], norm_fn=norm_fn, stride=2, padding=int(spconv_kernel_sizes[0]//2), indice_key='spconv2', conv_type='spconv'), + SparseBasicBlock(64, 64, norm_fn=norm_fn, indice_key='res2'), + SparseBasicBlock(64, 64, norm_fn=norm_fn, indice_key='res2'), + SparseBasicBlock(64, 64, norm_fn=norm_fn, indice_key='res2'), + SparseBasicBlock(64, 64, norm_fn=norm_fn, indice_key='res2'), + ) + + self.conv3 = spconv.SparseSequential( + # [800, 704] <- [400, 352] + block(64, 128, spconv_kernel_sizes[1], norm_fn=norm_fn, stride=2, padding=int(spconv_kernel_sizes[1]//2), indice_key='spconv3', conv_type='spconv'), + SparseBasicBlock(128, 128, norm_fn=norm_fn, indice_key='res3'), + SparseBasicBlock(128, 128, norm_fn=norm_fn, indice_key='res3'), + SparseBasicBlock(128, 128, norm_fn=norm_fn, indice_key='res3'), + SparseBasicBlock(128, 128, norm_fn=norm_fn, indice_key='res3'), + SparseBasicBlock(128, 128, norm_fn=norm_fn, indice_key='res3'), + SparseBasicBlock(128, 128, norm_fn=norm_fn, indice_key='res3'), + ) + + self.conv4 = spconv.SparseSequential( + # [400, 352] <- [200, 176] + block(128, 256, spconv_kernel_sizes[2], norm_fn=norm_fn, stride=2, padding=int(spconv_kernel_sizes[2]//2), indice_key='spconv4', conv_type='spconv'), + SparseBasicBlock(256, 256, norm_fn=norm_fn, indice_key='res4'), + SparseBasicBlock(256, 256, norm_fn=norm_fn, indice_key='res4'), + SparseBasicBlock(256, 256, norm_fn=norm_fn, indice_key='res4'), + ) + + self.conv5 = spconv.SparseSequential( + # [400, 352] <- [200, 176] + block(256, 256, spconv_kernel_sizes[3], norm_fn=norm_fn, stride=2, padding=int(spconv_kernel_sizes[3]//2), indice_key='spconv5', conv_type='spconv'), + SparseBasicBlock(256, 256, norm_fn=norm_fn, indice_key='res5'), + SparseBasicBlock(256, 256, norm_fn=norm_fn, indice_key='res5'), + SparseBasicBlock(256, 256, norm_fn=norm_fn, indice_key='res5'), + ) + + self.conv6 = spconv.SparseSequential( + # [400, 352] <- [200, 176] + block(256, 256, spconv_kernel_sizes[3], norm_fn=norm_fn, stride=2, padding=int(spconv_kernel_sizes[3]//2), indice_key='spconv6', conv_type='spconv'), + SparseBasicBlock(256, 256, norm_fn=norm_fn, indice_key='res6'), + SparseBasicBlock(256, 256, norm_fn=norm_fn, indice_key='res6'), + SparseBasicBlock(256, 256, norm_fn=norm_fn, indice_key='res6'), + ) + + self.conv_out = spconv.SparseSequential( + # [200, 150, 5] -> [200, 150, 2] + spconv.SparseConv2d(256, 256, 3, stride=1, padding=1, bias=False, indice_key='spconv_down2'), + norm_fn(256), + nn.ReLU(), + ) + + self.shared_conv = spconv.SparseSequential( + spconv.SubMConv2d(256, 256, 3, stride=1, padding=1, bias=True), + nn.BatchNorm1d(256), + nn.ReLU(True), + ) + + self.num_point_features = 256 + self.backbone_channels = { + 'x_conv1': 32, + 'x_conv2': 64, + 'x_conv3': 128, + 'x_conv4': 256, + 'x_conv5': 256 + } + self.forward_ret_dict = {} + + def bev_out(self, x_conv): + features_cat = x_conv.features + indices_cat = x_conv.indices + + indices_unique, _inv = torch.unique(indices_cat, dim=0, return_inverse=True) + features_unique = features_cat.new_zeros((indices_unique.shape[0], features_cat.shape[1])) + features_unique.index_add_(0, _inv, features_cat) + + x_out = spconv.SparseConvTensor( + features=features_unique, + indices=indices_unique, + spatial_shape=x_conv.spatial_shape, + batch_size=x_conv.batch_size + ) + return x_out + + def forward(self, batch_dict): + pillar_features, pillar_coords = batch_dict['pillar_features'], batch_dict['pillar_coords'] + batch_size = batch_dict['batch_size'] + input_sp_tensor = spconv.SparseConvTensor( + features=pillar_features, + indices=pillar_coords.int(), + spatial_shape=self.sparse_shape, + batch_size=batch_size + ) + + x_conv1 = self.conv1(input_sp_tensor) + x_conv2 = self.conv2(x_conv1) + x_conv3 = self.conv3(x_conv2) + x_conv4 = self.conv4(x_conv3) + x_conv5 = self.conv5(x_conv4) + x_conv6 = self.conv6(x_conv5) + + x_conv5.indices[:, 1:] *= 2 + x_conv6.indices[:, 1:] *= 4 + x_conv4 = x_conv4.replace_feature(torch.cat([x_conv4.features, x_conv5.features, x_conv6.features])) + x_conv4.indices = torch.cat([x_conv4.indices, x_conv5.indices, x_conv6.indices]) + + out = self.bev_out(x_conv4) + + out = self.conv_out(out) + out = self.shared_conv(out) + + batch_dict.update({ + 'encoded_spconv_tensor': out, + 'encoded_spconv_tensor_stride': 8 + }) + batch_dict.update({ + 'multi_scale_2d_features': { + 'x_conv1': x_conv1, + 'x_conv2': x_conv2, + 'x_conv3': x_conv3, + 'x_conv4': x_conv4, + 'x_conv5': x_conv5, + } + }) + batch_dict.update({ + 'multi_scale_2d_strides': { + 'x_conv1': 1, + 'x_conv2': 2, + 'x_conv3': 4, + 'x_conv4': 8, + 'x_conv5': 16, + } + }) + + return batch_dict