A Text Recognition Dataset from Sahidic Coptic Ancient Manuscripts
Not on the map. Why?
Abstract (arXiv)
In this work, we target Handwritten Text Recognition (HTR) in low-resource scenarios, which arise from underrepresented languages, rare scripts, and degraded visual conditions typical of historical documents. We introduce SCAM (Sahidic Coptic Ancient Manuscripts), a new line-level dataset built from digitized ancient manuscripts written in the extinct Sahidic Coptic dialect. The dataset reflects a realistic and challenging setting, as it combines heterogeneous acquisition conditions across libraries with typical manuscript degradations such as ink fading, bleed-through, and material deterioration. In addition to visual complexity, SCAM poses significant linguistic challenges due to the scarcity of resources for Sahidic Coptic, its uncommon alphabet, and dialect-specific diacritics. To support research in low-resource HTR, we benchmark several state-of-the-art approaches based on different paradigms, highlighting their limitations and strengths in this setting. Our results underline the gap between current HTR performance on well-resourced modern scripts and historically grounded, low-resource scenarios, thus providing a reference point for future developments.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 27 of its 66 references that are in the map. First posted on arXiv 14 Jun 2026.
Builds on
27 of 66 references.
References · 27
- 2017Attention Is All You Needcited by 60,361
- 2015Spatial Transformer Networkscited by 1,429
- 2018Demystifying MMD GANscited by 1,293
- 2023The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)cited by 681
- 2015An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognitioncited by 520
- 2021TrOCR: Transformer-based Optical Character Recognition with Pre-trained Modelscited by 279
- 2020Pay Attention to What You Read: Non-recurrent Handwritten Text-Line Recognitioncited by 71
- 2020Sequence-to-Sequence Contrastive Learning for Text Recognitioncited by 70
- 2020GANwriting: Content-Conditioned Generation of Styled Handwritten Word Imagescited by 68
- 2019Evaluating Sequence-to-Sequence Models for Handwritten Text Recognitioncited by 58
- 2020OrigamiNet: Weakly-Supervised, Segmentation-Free, One-Step, Full Page Text Recognition by learning to unfoldcited by 53
- 2020End-to-end Handwritten Paragraph Text Recognition Using a Vertical Attention Networkcited by 50
15 more
- 2016Joint Line Segmentation and Transcription for End-to-End Handwritten Paragraph Recognitioncited by 46
- 2016Scan, Attend and Read: End-to-End Handwritten Paragraph Recognition with MDLSTM Attentioncited by 45
- 2013Dropout improves Recurrent Neural Networks for Handwriting Recognitioncited by 39
- 2018Handwriting Recognition in Low-resource Scripts using Adversarial Learningcited by 32
- 2018An Efficient End-to-End Neural Model for Handwritten Text Recognitioncited by 25
- 2024HTR-VT: Handwritten Text Recognition with Vision Transformercited by 24
- 2023HWD: A Novel Evaluation Score for Styled Handwritten Text Generationcited by 19
- 2021Rescoring Sequence-to-Sequence Models for Text Line Recognition with CTC-Prefixescited by 18
- 2022The LAM Dataset: A Novel Benchmark for Line-Level Handwritten Text Recognitioncited by 17
- 2019KuroNet: Pre-Modern Japanese Kuzushiji Character Recognition with Deep Learningcited by 13
- 2019Fully Convolutional Networks for Handwriting Recognitioncited by 10
- 2018Are 2D-LSTM really dead for offline text recognition?cited by 7
- 2019No Padding Please: Efficient Neural Handwriting Recognitioncited by 5
- 2024General Detection-based Text Line Recognitioncited by 4
- 2024Binarizing Documents by Leveraging both Space and Frequencycited by 3
Built on by
Citing papers per year: 2026: 1.
1 paper.