Abstract
Despite Bengali being the sixth most spoken language in the world, handwritten text recognition (HTR) systems for Bengali remain severely underdeveloped. The complexity of Bengali script--featuring conjuncts, diacritics, and highly variable handwriting styles--combined with a scarcity of annotated datasets makes this task particularly challenging. We present GraDeT-HTR, a resource-efficient Bengali handwritten text recognition system based on a Grapheme-aware Decoder-only Transformer architecture. To address the unique challenges of Bengali script, we augment the performance of a decoder-only transformer by integrating a grapheme-based tokenizer and demonstrate that it significantly improves recognition accuracy compared to conventional subword tokenizers. Our model is pretrained on large-scale synthetic data and fine-tuned on real human-annotated samples, achieving state-of-the-art performance on multiple benchmark datasets.
Abstract (translated)
尽管孟加拉语是世界上使用人数第六多的语言，但孟加拉语的手写文本识别（HTR）系统仍然严重缺乏发展。孟加拉语的复杂性——包括连字符、声调符号以及高度变化的手写风格——再加上标注数据集的匮乏，使得这一任务尤其具有挑战性。我们提出了GraDeT-HTR，这是一个基于图素感知解码器独奏变压器架构的资源高效型孟加拉语文本识别系统。为了应对孟加拉语特有的挑战，我们在一个解码器独奏的变压器中整合了基于图素的标记器，并证明与传统的子词标记器相比，这能显著提高识别准确性。我们的模型在大规模合成数据上进行了预训练，并通过真实的标注样本进行微调，在多个基准数据集上实现了最佳性能。
URL
https://arxiv.org/abs/2509.18081
