A Transformer Architecture for Online Gesture Recognition of Mathematical Expressions

TL;DR

This paper introduces a Transformer-based model for online handwritten gesture recognition that accurately constructs mathematical expression trees, demonstrating robustness and high accuracy on a new dataset with potential applications beyond handwriting recognition.

Contribution

It presents a novel Transformer architecture that encodes spatio-temporal gesture data to produce syntactically correct mathematical expression trees, including a new dataset and evaluation metric.

Findings

Achieved 94% normalized Levenshtein accuracy on expression tree predictions.

Successfully trained a small Transformer suitable for edge devices.

Demonstrated robustness to input ablation and unseen glyphs.

Abstract

The Transformer architecture is shown to provide a powerful framework as an end-to-end model for building expression trees from online handwritten gestures corresponding to glyph strokes. In particular, the attention mechanism was successfully used to encode, learn and enforce the underlying syntax of expressions creating latent representations that are correctly decoded to the exact mathematical expression tree, providing robustness to ablated inputs and unseen glyphs. For the first time, the encoder is fed with spatio-temporal data tokens potentially forming an infinitely large vocabulary, which finds applications beyond that of online gesture recognition. A new supervised dataset of online handwriting gestures is provided for training models on generic handwriting recognition tasks and a new metric is proposed for the evaluation of the syntactic correctness of the output expression trees. A small Transformer model suitable for edge inference was successfully trained to an average normalised Levenshtein accuracy of 94%, resulting in valid postfix RPN tree representation for 94% of predictions.