Lacuna Reconstruction: Self-supervised Pre-training for Low-Resource Historical Document Transcription
This addresses the challenge of transcribing historical documents with limited labeled data, which is incremental as it builds on existing self-supervised methods for visual language representation.
The paper tackled the problem of low-resource historical document transcription by developing a self-supervised pre-training approach, resulting in meaningful improvements in recognition accuracy with as few as 30 line image transcriptions for training on handwritten Islamicate manuscripts and early modern English printed documents.
We present a self-supervised pre-training approach for learning rich visual language representations for both handwritten and printed historical document transcription. After supervised fine-tuning of our pre-trained encoder representations for low-resource document transcription on two languages, (1) a heterogeneous set of handwritten Islamicate manuscript images and (2) early modern English printed documents, we show a meaningful improvement in recognition accuracy over the same supervised model trained from scratch with as few as 30 line image transcriptions for training. Our masked language model-style pre-training strategy, where the model is trained to be able to identify the true masked visual representation from distractors sampled from within the same line, encourages learning robust contextualized language representations invariant to scribal writing style and printing noise present across documents.