Texo: Formula Recognition within 20M Parameters

TL;DR

Texo is a lightweight, high-performance formula recognition model with only 20 million parameters, achieving state-of-the-art accuracy while enabling real-time and in-browser deployment.

Contribution

The paper introduces Texo, a minimalist formula recognition model that reduces size significantly while maintaining high accuracy through attentive design and transfer learning.

Findings

Achieves comparable performance to larger models.

Reduces model size by 80% and 65%.

Enables real-time and in-browser deployment.

Abstract

In this paper we present Texo, a minimalist yet highperformance formula recognition model that contains only 20 million parameters. By attentive design, distillation and transfer of the vocabulary and the tokenizer, Texo achieves comparable performance to state-of-the-art models such as UniMERNet-T and PPFormulaNet-S, while reducing the model size by 80% and 65%, respectively. This enables real-time inference on consumer-grade hardware and even in-browser deployment. We also developed a web application to demonstrate the model capabilities and facilitate its usage for end users.