用高效微调让大模型精准识别海量开源字体
Parameter-Efficient Fine-Tuning of DINOv2 for Large-Scale Font Classification
- 用LoRA技术仅训练1%参数，实现99%准确率
- 在394种字体上达到99.0%分类准确率，错误严重度比随机猜测低140倍
- 专为开源字体设计，适合字体识别与高效模型应用研究者
我们提出GoogleFontsBench，首个面向开源网络字体的公开基准，填补了现有基准仅覆盖商业字体的空白。该基准包含32个Google Fonts家族的394种字形变体，采用可复现的合成数据生成流程（每变体约575张图像，总计约226,000张），并引入基于排版学的评估指标SWER，按视觉误差严重度加权。我们在DINOv2 Vision Transformer基础上，对比六种微调策略建立基线。采用参数高效适配方法LoRA，仅训练模型8720万参数中的1%，即实现99.0%的top-1准确率，且错误严重度比随机猜测低140倍。我们已开源该基准、所有训练模型及完整训练流程。
原文摘要 · Abstract (English)
We introduce GoogleFontsBench, the first public benchmark for classifying open-source web fonts, addressing a gap left by existing benchmarks that cover only commercial typefaces. GoogleFontsBench comprises 394 font variants across 32 Google Fonts families, a reproducible synthetic data generation pipeline (~575 images per variant, ~226K total), and a typographically-grounded evaluation metric (SWER) that weights errors by visual severity. We establish baselines using six fine-tuning strategies on a DINOv2 Vision Transformer backbone. Parameter-efficient adaptation with LoRA achieves 99.0% top-1 accuracy while training only 1% of the model's 87.2M parameters, with errors 140x less severe than random guessing. We release the benchmark, all trained models, and the full training pipeline as open-source resources.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。