This is a text spotting model that simultaneously detects and recognizes text. The model detects symbol sequences separated by space and performs recognition without a dictionary. The model is built on top of the Mask-RCNN framework with additional attention-based text recognition head.
Symbols set is alphanumeric:
This model is 2D attention-based GRU decoder of text recognition head.
|Word spotting hmean ICDAR2015, without a dictionary||59.04%|
Hmean Word spotting is defined and measured according to the Incidental Scene Text (ICDAR2015) challenge.
encoder_outputs, shape: [1x(28*28)x256]. Encoded text recognition features.
prev_symbol, shape: [1x1]. Index in alphabet of the previously generated symbol.
prev_hidden, shape: [1x1x256]. Previous hidden state of GRU.
output, shape: [1x38]. Encoded text recognition features.
hidden, shape: [1x1x256]. Current hidden state of GRU.
[*] Other names and brands may be claimed as the property of others.