Parameter-Efficient Fine-Tuning of DINOv2 for Large-Scale Font Classification

2026-02-14

AI Summary

This work proposes an efficient font classification method based on the DINOv2 vision transformer for large-scale recognition across 394 font families. By leveraging Low-Rank Adaptation (LoRA), the approach fine-tunes fewer than 1% of the model’s parameters while incorporating a scalable synthetic data generation pipeline that includes augmentations such as random colors, alignment variations, line breaks, and Gaussian noise to significantly enhance generalization. The method achieves a Top-1 accuracy of approximately 86%, marking the first application of LoRA with DINOv2 for font classification and demonstrating both high efficiency and strong performance. The authors have open-sourced the trained models, dataset, and the complete training and deployment pipeline.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Neural Architectures and Foundation Models
• Natural Language Processing: (Large) Language Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

We present a font classification system capable of identifying 394 font families from rendered text images. Our approach fine-tunes a DINOv2 Vision Transformer using Low-Rank Adaptation (LoRA), achieving approximately 86% top-1 accuracy while training fewer than 1% of the model's 87.2M parameters. We introduce a synthetic dataset generation pipeline that renders Google Fonts at scale with diverse augmentations including randomized colors, alignment, line wrapping, and Gaussian noise, producing training images that generalize to real-world typographic samples. The model incorporates built-in preprocessing to ensure consistency between training and inference, and is deployed as a HuggingFace Inference Endpoint. We release the model, dataset, and full training pipeline as open-source resources.

Problem

• font classification
• parameter-efficient fine-tuning
• large-scale
• DINOv2
• generalization

Innovation

• Parameter-Efficient Fine-Tuning
• Low-Rank Adaptation (LoRA)
• DINOv2 Vision Transformer
• Synthetic Font Dataset
• Font Classification