改进TrOCR模型以识别提格雷尼亚文，解决跨文字系统识别难题。
Adapting TrOCR for Printed Tigrinya Text Recognition: Word-Aware Loss Weighting for Cross-Script Transfer Learning
- 扩展字节级分词器并引入词级损失加权机制
- 在5000张合成图像上实现0.22%错误率和97.2%准确率
- 适合关注非洲文字识别与迁移学习的研究者
基于Transformer的OCR模型在拉丁文和中日韩文字上表现优异，但在非洲音节文字系统中的应用仍受限。本文首次将TrOCR adapted to printed Tigrinya（使用吉兹字母），从预训练模型出发，将字节级BPE分词器扩展至覆盖230个吉兹字符，并引入词级损失加权机制，解决因沿用拉丁文分词习惯导致的词边界错误问题。未经调整的原始模型对吉兹文本无有效输出。经适配后，TrOCR-Printed在来自GLOCR数据集的5,000张合成图像测试集上达到0.22%字符错误率（CER）和97.20%精确匹配准确率。消融实验表明，词级损失加权是关键，相比仅扩展词汇量，可使CER降低两个数量级。整个流程可在单块8GB消费级显卡上于三小时内完成训练。所有代码、模型权重与评估脚本均公开发布。
原文摘要 · Abstract (English)
Transformer-based OCR models have shown strong performance on Latin and CJK scripts, but their application to African syllabic writing systems remains limited. We present the first adaptation of TrOCR for printed Tigrinya using the Ge'ez script. Starting from a pre-trained model, we extend the byte-level BPE tokenizer to cover 230 Ge'ez characters and introduce Word-Aware Loss Weighting to resolve systematic word-boundary failures that arise when applying Latin-centric BPE conventions to a new script. The unmodified model produces no usable output on Ge'ez text. After adaptation, the TrOCR-Printed variant achieves 0.22% Character Error Rate and 97.20% exact match accuracy on a held-out test set of 5,000 synthetic images from the GLOCR dataset. An ablation study confirms that Word-Aware Loss Weighting is the critical component, reducing CER by two orders of magnitude compared to vocabulary extension alone. The full pipeline trains in under three hours on a single 8 GB consumer GPU. All code, model weights, and evaluation scripts are publicly released.
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。