# Copyright (C) 2021-2024, Mindee. # This program is licensed under the Apache License 2.0. # See LICENSE or go to for full license details. from copy import deepcopy from typing import Any, Dict, List, Optional, Tuple import tensorflow as tf from tensorflow.keras import Model, layers from doctr.datasets import VOCABS from ...classification import vit_b, vit_s from ...utils.tensorflow import _bf16_to_float32, load_pretrained_params from .base import _ViTSTR, _ViTSTRPostProcessor __all__ = ["ViTSTR", "vitstr_small", "vitstr_base"] default_cfgs: Dict[str, Dict[str, Any]] = { "vitstr_small": { "mean": (0.694, 0.695, 0.693), "std": (0.299, 0.296, 0.301), "input_shape": (32, 128, 3), "vocab": VOCABS["french"], "url": "https://doctr-static.mindee.com/models?id=v0.6.0/vitstr_small-358fab2e.zip&src=0", }, "vitstr_base": { "mean": (0.694, 0.695, 0.693), "std": (0.299, 0.296, 0.301), "input_shape": (32, 128, 3), "vocab": VOCABS["french"], "url": "https://doctr-static.mindee.com/models?id=v0.6.0/vitstr_base-2889159a.zip&src=0", }, } class ViTSTR(_ViTSTR, Model): """Implements a ViTSTR architecture as described in `"Vision Transformer for Fast and Efficient Scene Text Recognition" `_. Args: ---- feature_extractor: the backbone serving as feature extractor vocab: vocabulary used for encoding embedding_units: number of embedding units max_length: maximum word length handled by the model dropout_prob: dropout probability for the encoder and decoder input_shape: input shape of the image exportable: onnx exportable returns only logits cfg: dictionary containing information about the model """ _children_names: List[str] = ["feat_extractor", "postprocessor"] def __init__( self, feature_extractor, vocab: str, embedding_units: int, max_length: int = 32, dropout_prob: float = 0.0, input_shape: Tuple[int, int, int] = (32, 128, 3), # different from paper exportable: bool = False, cfg: Optional[Dict[str, Any]] = None, ) -> None: super().__init__() self.vocab = vocab self.exportable = exportable self.cfg = cfg self.max_length = max_length + 2 # +2 for SOS and EOS self.feat_extractor = feature_extractor self.head = layers.Dense(len(self.vocab) + 1, name="head") # +1 for EOS self.postprocessor = ViTSTRPostProcessor(vocab=self.vocab) @staticmethod def compute_loss( model_output: tf.Tensor, gt: tf.Tensor, seq_len: List[int], ) -> tf.Tensor: """Compute categorical cross-entropy loss for the model. Sequences are masked after the EOS character. Args: ---- model_output: predicted logits of the model gt: the encoded tensor with gt labels seq_len: lengths of each gt word inside the batch Returns: ------- The loss of the model on the batch """ # Input length : number of steps input_len = tf.shape(model_output)[1] # Add one for additional token (sos disappear in shift!) seq_len = tf.cast(seq_len, tf.int32) + 1 # One-hot gt labels oh_gt = tf.one_hot(gt, depth=model_output.shape[2]) # Compute loss: don't forget to shift gt! Otherwise the model learns to output the gt[t-1]! # The "masked" first gt char is . cce = tf.nn.softmax_cross_entropy_with_logits(oh_gt[:, 1:, :], model_output) # Compute mask mask_values = tf.zeros_like(cce) mask_2d = tf.sequence_mask(seq_len, input_len) masked_loss = tf.where(mask_2d, cce, mask_values) ce_loss = tf.math.divide(tf.reduce_sum(masked_loss, axis=1), tf.cast(seq_len, model_output.dtype)) return tf.expand_dims(ce_loss, axis=1) def call( self, x: tf.Tensor, target: Optional[List[str]] = None, return_model_output: bool = False, return_preds: bool = False, **kwargs: Any, ) -> Dict[str, Any]: features = self.feat_extractor(x, **kwargs) # (batch_size, patches_seqlen, d_model) if target is not None: gt, seq_len = self.build_target(target) seq_len = tf.cast(seq_len, tf.int32) if kwargs.get("training", False) and target is None: raise ValueError("Need to provide labels during training") features = features[:, : self.max_length] # (batch_size, max_length, d_model) B, N, E = features.shape features = tf.reshape(features, (B * N, E)) logits = tf.reshape( self.head(features, **kwargs), (B, N, len(self.vocab) + 1) ) # (batch_size, max_length, vocab + 1) decoded_features = _bf16_to_float32(logits[:, 1:]) # remove cls_token out: Dict[str, tf.Tensor] = {} if self.exportable: out["logits"] = decoded_features return out if return_model_output: out["out_map"] = decoded_features if target is None or return_preds: # Post-process boxes out["preds"] = self.postprocessor(decoded_features) if target is not None: out["loss"] = self.compute_loss(decoded_features, gt, seq_len) return out class ViTSTRPostProcessor(_ViTSTRPostProcessor): """Post processor for ViTSTR architecture Args: ---- vocab: string containing the ordered sequence of supported characters """ def __call__( self, logits: tf.Tensor, ) -> List[Tuple[str, float]]: # compute pred with argmax for attention models out_idxs = tf.math.argmax(logits, axis=2) preds_prob = tf.math.reduce_max(tf.nn.softmax(logits, axis=-1), axis=-1) # decode raw output of the model with tf_label_to_idx out_idxs = tf.cast(out_idxs, dtype="int32") embedding = tf.constant(self._embedding, dtype=tf.string) decoded_strings_pred = tf.strings.reduce_join(inputs=tf.nn.embedding_lookup(embedding, out_idxs), axis=-1) decoded_strings_pred = tf.strings.split(decoded_strings_pred, "") decoded_strings_pred = tf.sparse.to_dense(decoded_strings_pred.to_sparse(), default_value="not valid")[:, 0] word_values = [word.decode() for word in decoded_strings_pred.numpy().tolist()] # compute probabilties for each word up to the EOS token probs = [ preds_prob[i, : len(word)].numpy().clip(0, 1).mean().item() if word else 0.0 for i, word in enumerate(word_values) ] return list(zip(word_values, probs)) def _vitstr( arch: str, pretrained: bool, backbone_fn, input_shape: Optional[Tuple[int, int, int]] = None, **kwargs: Any, ) -> ViTSTR: # Patch the config _cfg = deepcopy(default_cfgs[arch]) _cfg["input_shape"] = input_shape or _cfg["input_shape"] _cfg["vocab"] = kwargs.get("vocab", _cfg["vocab"]) patch_size = kwargs.get("patch_size", (4, 8)) kwargs["vocab"] = _cfg["vocab"] # Feature extractor feat_extractor = backbone_fn( # NOTE: we don't use a pretrained backbone for non-rectangular patches to avoid the pos embed mismatch pretrained=False, input_shape=_cfg["input_shape"], patch_size=patch_size, include_top=False, ) kwargs.pop("patch_size", None) kwargs.pop("pretrained_backbone", None) # Build the model model = ViTSTR(feat_extractor, cfg=_cfg, **kwargs) # Load pretrained parameters if pretrained: load_pretrained_params(model, default_cfgs[arch]["url"]) return model def vitstr_small(pretrained: bool = False, **kwargs: Any) -> ViTSTR: """ViTSTR-Small as described in `"Vision Transformer for Fast and Efficient Scene Text Recognition" `_. >>> import tensorflow as tf >>> from doctr.models import vitstr_small >>> model = vitstr_small(pretrained=False) >>> input_tensor = tf.random.uniform(shape=[1, 32, 128, 3], maxval=1, dtype=tf.float32) >>> out = model(input_tensor) Args: ---- pretrained (bool): If True, returns a model pre-trained on our text recognition dataset **kwargs: keyword arguments of the ViTSTR architecture Returns: ------- text recognition architecture """ return _vitstr( "vitstr_small", pretrained, vit_s, embedding_units=384, patch_size=(4, 8), **kwargs, ) def vitstr_base(pretrained: bool = False, **kwargs: Any) -> ViTSTR: """ViTSTR-Base as described in `"Vision Transformer for Fast and Efficient Scene Text Recognition" `_. >>> import tensorflow as tf >>> from doctr.models import vitstr_base >>> model = vitstr_base(pretrained=False) >>> input_tensor = tf.random.uniform(shape=[1, 32, 128, 3], maxval=1, dtype=tf.float32) >>> out = model(input_tensor) Args: ---- pretrained (bool): If True, returns a model pre-trained on our text recognition dataset **kwargs: keyword arguments of the ViTSTR architecture Returns: ------- text recognition architecture """ return _vitstr( "vitstr_base", pretrained, vit_b, embedding_units=768, patch_size=(4, 8), **kwargs, )