Builds a VIPBlock performing global multi-head self-attention. Args: embed_dim: embedding dimension. depth: number of attention b

Builds a VIPBlock performing local (cross-shaped) window attention. Args: embed_dim: embedding dimension. depth: number of attent

Builds a VIPBlock performing mixed local+global attention. Args: embed_dim: embedding dimension. depth: number of attention block

Convert a bounding box to a polygon Args: bbox: a bounding box Returns: a polygon

Compute boxes from a bitmap/pred_map: find connected components then filter boxes Args: pred: Pred map from differentiable linkne

Compute boxes from a bitmap/pred_map: find connected components then filter boxes Args: pred: Pred map from differentiable linkne

Compute boxes from a bitmap/pred_map: find connected components then filter boxes Args: pred: Pred map from differentiable binari

Encode a list of gts sequences into a np array and gives the corresponding* sequence lengths. Args: gts: list of ground-t

Converts target to relative coordinates Args: img: tf.Tensor or torch.Tensor representing the image target: target to convert to

CRNN with a MobileNet V3 Large backbone as described in `"An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Appl

CRNN with a MobileNet V3 Small backbone as described in `"An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Appl

CRNN with a VGG-16 backbone as described in `"An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to S

DBNet as described in `"Real-time Scene Text Detection with Differentiable Binarization" <https://arxiv.org/pdf/1911.08947.pdf>`_, using a MobileN

DBNet as described in `"Real-time Scene Text Detection with Differentiable Binarization" <https://arxiv.org/pdf/1911.08947.pdf>`_, using a ResNet-

DBNet as described in `"Real-time Scene Text Detection with Differentiable Binarization" <https://arxiv.org/pdf/1911.08947.pdf>`_, using a ResNet-

Draw an array of relative straight boxes on an image Args: boxes: array of relative boxes, of shape (*, 4) image: np array, float

Given a predefined mapping, encode the string to a sequence of numbers Args: input_string: string to encode vocab: vocabulary (st

FAST as described in `"FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation" <https://arxiv.org/pdf/2111.02394.pdf>

FAST as described in `"FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation" <https://arxiv.org/pdf/2111.02394.pdf>

FAST as described in `"FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation" <https://arxiv.org/pdf/2111.02394.pdf>

Forward pass Args: x: embeddings (batch, max_len, d_model) Returns: positional embeddings (batch, max_len, d

Call function for training Args: x: images target: list of str labels return_model_output: if True, retur

Forward pass for VIPBlock. Args: x: input tensor (B, H, W, C) Returns: Transformed tensor