1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
|
"""A DETR style transfomers but for text recognition."""
from typing import Dict, Optional, Tuple
from einops import rearrange
import torch
from torch import nn
from torch import Tensor
from text_recognizer.networks.transformer import PositionalEncoding, Transformer
from text_recognizer.networks.util import configure_backbone
class CNNTransformer(nn.Module):
"""CNN+Transfomer for image to sequence prediction, sort of based on the ideas from DETR."""
def __init__(
self,
num_encoder_layers: int,
num_decoder_layers: int,
hidden_dim: int,
vocab_size: int,
num_heads: int,
adaptive_pool_dim: Tuple,
expansion_dim: int,
dropout_rate: float,
trg_pad_index: int,
backbone: str,
out_channels: int,
max_len: int,
backbone_args: Optional[Dict] = None,
activation: str = "gelu",
) -> None:
super().__init__()
self.trg_pad_index = trg_pad_index
self.backbone = configure_backbone(backbone, backbone_args)
self.character_embedding = nn.Embedding(vocab_size, hidden_dim)
# self.conv = nn.Conv2d(out_channels, max_len, kernel_size=1)
self.position_encoding = PositionalEncoding(hidden_dim, dropout_rate)
self.row_embed = nn.Parameter(torch.rand(max_len, max_len // 2))
self.col_embed = nn.Parameter(torch.rand(max_len, max_len // 2))
self.adaptive_pool = (
nn.AdaptiveAvgPool2d((adaptive_pool_dim)) if adaptive_pool_dim else None
)
self.transformer = Transformer(
num_encoder_layers,
num_decoder_layers,
hidden_dim,
num_heads,
expansion_dim,
dropout_rate,
activation,
)
self.head = nn.Sequential(nn.Linear(hidden_dim, vocab_size),)
def _create_trg_mask(self, trg: Tensor) -> Tensor:
# Move this outside the transformer.
trg_pad_mask = (trg != self.trg_pad_index)[:, None, None]
trg_len = trg.shape[1]
trg_sub_mask = torch.tril(
torch.ones((trg_len, trg_len), device=trg.device)
).bool()
trg_mask = trg_pad_mask & trg_sub_mask
return trg_mask
def encoder(self, src: Tensor) -> Tensor:
"""Forward pass with the encoder of the transformer."""
return self.transformer.encoder(src)
def decoder(self, trg: Tensor, memory: Tensor, trg_mask: Tensor) -> Tensor:
"""Forward pass with the decoder of the transformer + classification head."""
return self.head(
self.transformer.decoder(trg=trg, memory=memory, trg_mask=trg_mask)
)
def preprocess_input(self, src: Tensor) -> Tensor:
"""Encodes src with a backbone network and a positional encoding.
Args:
src (Tensor): Input tensor.
Returns:
Tensor: A input src to the transformer.
"""
# If batch dimenstion is missing, it needs to be added.
if len(src.shape) < 4:
src = src[(None,) * (4 - len(src.shape))]
src = self.backbone(src)
# src = self.conv(src)
if self.adaptive_pool is not None:
src = self.adaptive_pool(src)
H, W = src.shape[-2:]
src = rearrange(src, "b t h w -> b t (h w)")
# construct positional encodings
pos = torch.cat(
[
self.col_embed[:W].unsqueeze(0).repeat(H, 1, 1),
self.row_embed[:H].unsqueeze(1).repeat(1, W, 1),
],
dim=-1,
).unsqueeze(0)
pos = rearrange(pos, "b h w l -> b l (h w)")
src = pos + 0.1 * src
return src
def preprocess_target(self, trg: Tensor) -> Tuple[Tensor, Tensor]:
"""Encodes target tensor with embedding and postion.
Args:
trg (Tensor): Target tensor.
Returns:
Tuple[Tensor, Tensor]: Encoded target tensor and target mask.
"""
trg = self.character_embedding(trg.long())
trg = self.position_encoding(trg)
return trg
def forward(self, x: Tensor, trg: Optional[Tensor] = None) -> Tensor:
"""Forward pass with CNN transfomer."""
h = self.preprocess_input(x)
trg_mask = self._create_trg_mask(trg)
trg = self.preprocess_target(trg)
out = self.transformer(h, trg, trg_mask=trg_mask)
logits = self.head(out)
return logits
|
