通过渐进式削弱文本标签影响，提升场景文字识别精度。
TEACH: Text Encoding as Curriculum Hints for Scene Text Recognition
- 将真实文本作为辅助输入，逐步减少其影响以模拟教学过程
- 在多个基准上显著提升识别准确率，尤其在复杂条件下表现更优
- 无需预训练、无推理开销，适配各类主流识别模型
场景文字识别（STR）因视觉外观复杂且语义先验有限而仍具挑战。本文提出TEACH，一种新颖的训练范式：将真实文本作为辅助输入注入模型，并在训练过程中逐步减弱其影响。通过将目标标签编码至嵌入空间并应用损失感知掩码，TEACH模拟了课程学习过程，引导模型从依赖标签的学习过渡到完全基于视觉的识别。与基于语言模型的方法不同，TEACH无需外部预训练，不引入推理开销，且具有模型无关性，可无缝集成至现有编码器-解码器框架中。在多个公开基准上的大量实验表明，使用TEACH训练的模型在各类条件下均实现一致的性能提升，验证了其鲁棒性和通用性。
原文摘要 · Abstract (English)
Scene Text Recognition (STR) remains a challenging task due to complex visual appearances and limited semantic priors. We propose TEACH, a novel training paradigm that injects ground-truth text into the model as auxiliary input and progressively reduces its influence during training. By encoding target labels into the embedding space and applying loss-aware masking, TEACH simulates a curriculum learning process that guides the model from label-dependent learning to fully visual recognition. Unlike language model-based approaches, TEACH requires no external pretraining and introduces no inference overhead. It is model-agnostic and can be seamlessly integrated into existing encoder-decoder frameworks. Extensive experiments across multiple public benchmarks show that models trained with TEACH achieve consistently improved accuracy, especially under challenging conditions, validating its robustness and general applicability.
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。