243 lines
9.2 KiB
Python
243 lines
9.2 KiB
Python
import torch
|
|
import torch.nn as nn
|
|
import torch.nn.functional as F
|
|
from .torch_utils import BaseConv, linear_init_
|
|
from .block import RepC3
|
|
from .detr_torch import TransformerEncoder
|
|
from .build_components import TRANSFORMER_ENCODER_REGISTRY
|
|
|
|
__all__ = ['ELAEncoder']
|
|
|
|
|
|
class TransformerLayer(nn.Module):
|
|
def __init__(self,
|
|
d_model=256,
|
|
nhead=8,
|
|
dim_feedforward=1024,
|
|
dropout=0.,
|
|
activation="relu",
|
|
attn_dropout=None,
|
|
act_dropout=None,
|
|
normalize_before=False):
|
|
super(TransformerLayer, self).__init__()
|
|
attn_dropout = dropout if attn_dropout is None else attn_dropout
|
|
act_dropout = dropout if act_dropout is None else act_dropout
|
|
self.normalize_before = normalize_before
|
|
|
|
self.self_attn = torch.nn.MultiheadAttention(d_model, nhead, attn_dropout)
|
|
# Implementation of Feedforward model
|
|
self.linear1 = nn.Linear(d_model, dim_feedforward)
|
|
self.dropout = nn.Dropout(act_dropout)
|
|
self.linear2 = nn.Linear(dim_feedforward, d_model)
|
|
|
|
self.norm1 = nn.LayerNorm(d_model)
|
|
self.norm2 = nn.LayerNorm(d_model)
|
|
self.dropout1 = nn.Dropout(dropout)
|
|
self.dropout2 = nn.Dropout(dropout)
|
|
self.activation = getattr(F, activation)
|
|
self._reset_parameters()
|
|
|
|
def _reset_parameters(self):
|
|
linear_init_(self.linear1)
|
|
linear_init_(self.linear2)
|
|
|
|
@staticmethod
|
|
def with_pos_embed(tensor, pos_embed):
|
|
return tensor if pos_embed is None else tensor + pos_embed
|
|
|
|
def forward(self, src, src_mask=None, pos_embed=None):
|
|
residual = src
|
|
if self.normalize_before:
|
|
src = self.norm1(src)
|
|
q = k = self.with_pos_embed(src, pos_embed)
|
|
src = self.self_attn(q, k, value=src, attn_mask=src_mask)
|
|
#print(src[1].shape, src[0].shape)
|
|
src = src[0]
|
|
src = residual + self.dropout1(src)
|
|
if not self.normalize_before:
|
|
src = self.norm1(src)
|
|
|
|
residual = src
|
|
if self.normalize_before:
|
|
src = self.norm2(src)
|
|
src = self.linear2(self.dropout(self.activation(self.linear1(src))))
|
|
src = residual + self.dropout2(src)
|
|
if not self.normalize_before:
|
|
src = self.norm2(src)
|
|
return src
|
|
|
|
|
|
@TRANSFORMER_ENCODER_REGISTRY.register()
|
|
class ELAEncoder(nn.Module):
|
|
# __shared__ = ['depth_mult', 'act', 'trt', 'eval_size']
|
|
# __inject__ = ['encoder_layer']
|
|
|
|
def __init__(self,
|
|
in_channels=[128, 256, 512],
|
|
feat_strides=[8, 16, 32],
|
|
hidden_dim=256,
|
|
use_encoder_idx=[2],
|
|
num_encoder_layers=1,
|
|
encoder_layer='TransformerLayer',
|
|
pe_temperature=10000,
|
|
expansion=1.0,
|
|
depth_mult=1.0,
|
|
act='silu',
|
|
trt=False,
|
|
dim_feedforward=1024,
|
|
eval_size=None):
|
|
super(ELAEncoder, self).__init__()
|
|
self.in_channels = in_channels
|
|
self.feat_strides = feat_strides
|
|
self.hidden_dim = hidden_dim
|
|
self.use_encoder_idx = use_encoder_idx
|
|
self.num_encoder_layers = num_encoder_layers
|
|
self.pe_temperature = pe_temperature
|
|
self.eval_size = eval_size
|
|
|
|
self.encoder_layer = TransformerLayer(dim_feedforward=dim_feedforward)
|
|
|
|
# channel projection
|
|
self.input_proj = nn.ModuleList()
|
|
for in_channel in self.in_channels:
|
|
self.input_proj.append(
|
|
nn.Sequential(
|
|
nn.Conv2d(
|
|
in_channel, hidden_dim, kernel_size=(1, 1), bias=False),
|
|
nn.BatchNorm2d(
|
|
hidden_dim)))
|
|
# encoder transformer
|
|
self.encoder = nn.ModuleList([
|
|
TransformerEncoder(self.encoder_layer, num_encoder_layers)
|
|
for _ in range(len(use_encoder_idx))
|
|
])
|
|
|
|
# act = get_act_fn(
|
|
# act, trt=trt) if act is None or isinstance(act,
|
|
# (str, dict)) else act
|
|
# top-down fpn
|
|
self.lateral_convs = nn.ModuleList()
|
|
self.fpn_blocks = nn.ModuleList()
|
|
for idx in range(len(self.in_channels) - 1, 0, -1):
|
|
self.lateral_convs.append(
|
|
BaseConv(
|
|
hidden_dim, hidden_dim, 1, 1, act=act))
|
|
self.fpn_blocks.append(
|
|
RepC3(
|
|
hidden_dim * 2,
|
|
hidden_dim,
|
|
round(3 * depth_mult),
|
|
e=1.0))
|
|
|
|
# bottom-up pan
|
|
self.downsample_convs = nn.ModuleList()
|
|
self.pan_blocks = nn.ModuleList()
|
|
for idx in range(len(self.in_channels) - 1):
|
|
self.downsample_convs.append(
|
|
BaseConv(
|
|
hidden_dim, hidden_dim, 3, stride=2, act=act))
|
|
self.pan_blocks.append(
|
|
RepC3(
|
|
hidden_dim * 2,
|
|
hidden_dim,
|
|
round(3 * depth_mult),
|
|
e=1.0))
|
|
|
|
# self._reset_parameters()
|
|
#
|
|
# def _reset_parameters(self):
|
|
# if self.eval_size:
|
|
# for idx in self.use_encoder_idx:
|
|
# stride = self.feat_strides[idx]
|
|
# pos_embed = self.build_2d_sincos_position_embedding(
|
|
# self.eval_size[1] // stride, self.eval_size[0] // stride,
|
|
# self.hidden_dim, self.pe_temperature)
|
|
# setattr(self, f'pos_embed{idx}', pos_embed)
|
|
|
|
@staticmethod
|
|
def build_2d_sincos_position_embedding(w,
|
|
h,
|
|
embed_dim=256,
|
|
temperature=10000.):
|
|
grid_w = torch.arange(int(w), dtype=torch.float32)
|
|
grid_h = torch.arange(int(h), dtype=torch.float32)
|
|
grid_w, grid_h = torch.meshgrid(grid_w, grid_h)
|
|
assert embed_dim % 4 == 0, \
|
|
'Embed dimension must be divisible by 4 for 2D sin-cos position embedding'
|
|
pos_dim = embed_dim // 4
|
|
omega = torch.arange(pos_dim, dtype=torch.float32) / pos_dim
|
|
omega = 1. / (temperature**omega)
|
|
|
|
out_w = grid_w.flatten()[..., None] @omega[None]
|
|
out_h = grid_h.flatten()[..., None] @omega[None]
|
|
|
|
return torch.cat(
|
|
[
|
|
torch.sin(out_w), torch.cos(out_w), torch.sin(out_h),
|
|
torch.cos(out_h)
|
|
],
|
|
dim=1)[None, :, :]
|
|
@classmethod
|
|
def from_config(cls, cfg):
|
|
enc_cfg = cfg.MODEL.ELAEncoder
|
|
return {
|
|
'in_channels': enc_cfg.in_channels,
|
|
'feat_strides': enc_cfg.feat_strides,
|
|
'hidden_dim': enc_cfg.hidden_dim,
|
|
'use_encoder_idx': enc_cfg.use_encoder_idx,
|
|
'num_encoder_layers': enc_cfg.num_encoder_layers,
|
|
'encoder_layer': enc_cfg.encoder_layer,
|
|
'pe_temperature': enc_cfg.pe_temperature,
|
|
'expansion': enc_cfg.expansion,
|
|
'depth_mult': enc_cfg.depth_mult,
|
|
'act': enc_cfg.act,
|
|
'eval_size': enc_cfg.eval_size,
|
|
'dim_feedforward': enc_cfg.dim_feedforward
|
|
}
|
|
|
|
def forward(self, feats, for_mot=False):
|
|
assert len(feats) == len(self.in_channels)
|
|
# get projection features
|
|
proj_feats = [self.input_proj[i](feat) for i, feat in enumerate(feats)]
|
|
# encoder
|
|
if self.num_encoder_layers > 0:
|
|
for i, enc_ind in enumerate(self.use_encoder_idx):
|
|
h, w = proj_feats[enc_ind].shape[2:]
|
|
# flatten [B, C, H, W] to [B, HxW, C]
|
|
src_flatten = proj_feats[enc_ind].flatten(start_dim=2).transpose(1, 2)
|
|
if self.training or self.eval_size is None:
|
|
pos_embed = self.build_2d_sincos_position_embedding(
|
|
w, h, self.hidden_dim, self.pe_temperature).to(src_flatten.device)
|
|
else:
|
|
pos_embed = getattr(self, f'pos_embed{enc_ind}', None)
|
|
memory = self.encoder[i](src_flatten, pos_embed=pos_embed)
|
|
proj_feats[enc_ind] = memory.transpose(1, 2).reshape((-1, self.hidden_dim, h, w))
|
|
|
|
# top-down fpn
|
|
inner_outs = [proj_feats[-1]]
|
|
for idx in range(len(self.in_channels) - 1, 0, -1):
|
|
feat_heigh = inner_outs[0]
|
|
feat_low = proj_feats[idx - 1]
|
|
feat_heigh = self.lateral_convs[len(self.in_channels) - 1 - idx](
|
|
feat_heigh)
|
|
inner_outs[0] = feat_heigh
|
|
|
|
upsample_feat = F.interpolate(
|
|
feat_heigh, scale_factor=2., mode="nearest")
|
|
inner_out = self.fpn_blocks[len(self.in_channels) - 1 - idx](
|
|
torch.cat(
|
|
[upsample_feat, feat_low], dim=1))
|
|
inner_outs.insert(0, inner_out)
|
|
|
|
# bottom-up pan
|
|
outs = [inner_outs[0]]
|
|
for idx in range(len(self.in_channels) - 1):
|
|
feat_low = outs[-1]
|
|
feat_height = inner_outs[idx + 1]
|
|
downsample_feat = self.downsample_convs[idx](feat_low)
|
|
out = self.pan_blocks[idx](torch.cat(
|
|
[downsample_feat, feat_height], dim=1))
|
|
outs.append(out)
|
|
|
|
return outs
|