Description We introduce LPR-MNIST, a synthetic dataset replicating the key aspects of syntax evolution on vehicle license plates, as explored in the paper Relaxed syntax modeling in Transformers for future-proof license plate recognition. LPR-MNIST is a collection of 100,000 synthetic image-text pairs, generated by concatenating 5 black-and-white digits from MNIST, each…
The Synthetic Text line Image Generator (STIG) is a software dedicated to handwritten textline generation. It is available for free under conditions (see below). Description The Synthetic Text line Image Generator is a Python code dedicated to the generation of synthetic handwritten textline images. The software can generate textlines looking…
The Modular Light Transformer (MLT) is a software dedicated to text recognition. It is available for free under conditions (see below). Description The Modular Light Transformer is a software dedicated to text recognition. It is based on a lightweight Transformer-based neural network dealing with text-line images. The proposed software is…
The aim of this thesis will be to look at self-supervised learning in order to take advantage of unsu-pervised examples for handwriting recognition. In recent years, a great deal of work has focused onself-supervised approaches for network pre-training. Starting with a pretext task that does not requiremanual annotation, the model…