# Copyright 2021 biubug6 . All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the 'License');
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an 'AS IS' BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# This code is modified by Zilliz.
#adapted from https://github.com/biubug6/Pytorch_Retinaface
import torch
from torch import nn
[docs]class ClassHead(nn.Module):
"""
ClassHead
RetinaFace head for classification branch.
Args:
inchannels (`int`):
number of input channels.
num_anchors (`int`):
number of anchors.
"""
[docs] def __init__(self,inchannels: int=512,num_anchors: int=3):
super().__init__()
self.num_anchors = num_anchors
self.conv1x1 = nn.Conv2d(inchannels,self.num_anchors*2,kernel_size=(1,1),stride=1,padding=0)
[docs] def forward(self,x: torch.FloatTensor):
out = self.conv1x1(x)
out = out.permute(0,2,3,1).contiguous()
return out.view(out.shape[0], -1, 2)
[docs]class BboxHead(nn.Module):
"""
BboxHead
RetinaFace head for bounding box branch.
Args:
inchannels (`int`):
number of input channels.
num_anchors (`int`):
number of anchors.
"""
[docs] def __init__(self,inchannels: int=512,num_anchors: int=3):
super().__init__()
self.conv1x1 = nn.Conv2d(inchannels,num_anchors*4,kernel_size=(1,1),stride=1,padding=0)
[docs] def forward(self,x: torch.FloatTensor):
out = self.conv1x1(x)
out = out.permute(0,2,3,1).contiguous()
return out.view(out.shape[0], -1, 4)
[docs]class LandmarkHead(nn.Module):
"""
LandmarkHead
RetinaFace head for landmark branch.
inchannels (`int`):
number of input channels.
num_anchors (`int`):
number of anchors.
"""
[docs] def __init__(self,inchannels: int=512,num_anchors: int=3):
super().__init__()
self.conv1x1 = nn.Conv2d(inchannels,num_anchors*10,kernel_size=(1,1),stride=1,padding=0)
[docs] def forward(self,x: torch.FloatTensor):
out = self.conv1x1(x)
out = out.permute(0,2,3,1).contiguous()
return out.view(out.shape[0], -1, 10)
