Move greedy decoder

2 files changed, 53 insertions, 0 deletions

diff --git a/text_recognizer/decoder/__init__.py b/text_recognizer/decoder/__init__.py
new file mode 100644
index 0000000..e69de29
--- /dev/null
+++ b/text_recognizer/decoder/__init__.py
diff --git a/text_recognizer/decoder/greedy_decoder.py b/text_recognizer/decoder/greedy_decoder.py
new file mode 100644
index 0000000..8d55a02
--- /dev/null
+++ b/text_recognizer/decoder/greedy_decoder.py
@@ -0,0 +1,53 @@
+"""Greedy decoder."""
+from typing import Type
+from text_recognizer.data.tokenizer import Tokenizer
+import torch
+from torch import nn, Tensor
+
+
+class GreedyDecoder:
+    def __init__(
+        self,
+        network: Type[nn.Module],
+        tokenizer: Tokenizer,
+        max_output_len: int = 682,
+    ) -> None:
+        self.network = network
+        self.start_index = tokenizer.start_index
+        self.end_index = tokenizer.end_index
+        self.pad_index = tokenizer.pad_index
+        self.max_output_len = max_output_len
+
+    def __call__(self, x: Tensor) -> Tensor:
+        bsz = x.shape[0]
+
+        # Encode image(s) to latent vectors.
+        img_features = self.network.encode(x)
+
+        # Create a placeholder matrix for storing outputs from the network
+        indecies = (
+            torch.ones((bsz, self.max_output_len), dtype=torch.long, device=x.device)
+            * self.pad_index
+        )
+        indecies[:, 0] = self.start_index
+
+        for i in range(1, self.max_output_len):
+            tokens = indecies[:, :i]  # (B, Sy)
+            logits = self.network.decode(tokens, img_features)  # [ B, N, C ]
+            indecies_ = logits.argmax(dim=2)  # [ B, N ]
+            indecies[:, i] = indecies_[:, -1]
+
+            # Early stopping of prediction loop if token is end or padding token.
+            if (
+                (indecies[:, i] == self.end_index) | (indecies[:, i] == self.pad_index)
+            ).all():
+                break
+
+        # Set all tokens after end token to pad token.
+        for i in range(1, self.max_output_len):
+            idx = (indecies[:, i - 1] == self.end_index) | (
+                indecies[:, i - 1] == self.pad_index
+            )
+            indecies[idx, i] = self.pad_index
+
+        return indecies