NENET: An Edge Learnable Network for Link Prediction in Scene Text
This addresses the challenge of connecting spatially separated and arbitrarily oriented characters in scene text detection, which is an incremental improvement over existing methods.
The paper tackles the problem of linking adjacent characters for scene text detection by proposing a novel Graph Neural Network (GNN) architecture that learns both node and edge features, achieving top results on the SynthText dataset compared to state-of-the-art methods.
Text detection in scenes based on deep neural networks have shown promising results. Instead of using word bounding box regression, recent state-of-the-art methods have started focusing on character bounding box and pixel-level prediction. This necessitates the need to link adjacent characters, which we propose in this paper using a novel Graph Neural Network (GNN) architecture that allows us to learn both node and edge features as opposed to only the node features under the typical GNN. The main advantage of using GNN for link prediction lies in its ability to connect characters which are spatially separated and have an arbitrary orientation. We show our concept on the well known SynthText dataset, achieving top results as compared to state-of-the-art methods.