تقرير
Improving Document Image Understanding with Reinforcement Finetuning
العنوان: | Improving Document Image Understanding with Reinforcement Finetuning |
---|---|
المؤلفون: | Nguyen, Bao-Sinh, Le, Dung Tien, Vu, Hieu M., Nguyen, Tuan Anh D., Nguyen, Minh-Tien, Le, Hung |
سنة النشر: | 2022 |
المجموعة: | Computer Science |
مصطلحات موضوعية: | Computer Science - Information Retrieval, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning |
الوصف: | Successful Artificial Intelligence systems often require numerous labeled data to extract information from document images. In this paper, we investigate the problem of improving the performance of Artificial Intelligence systems in understanding document images, especially in cases where training data is limited. We address the problem by proposing a novel finetuning method using reinforcement learning. Our approach treats the Information Extraction model as a policy network and uses policy gradient training to update the model to maximize combined reward functions that complement the traditional cross-entropy losses. Our experiments on four datasets using labels and expert feedback demonstrate that our finetuning mechanism consistently improves the performance of a state-of-the-art information extractor, especially in the small training data regime. Comment: Accepted to ICONIP 2022 |
نوع الوثيقة: | Working Paper |
URL الوصول: | http://arxiv.org/abs/2209.12561 |
رقم الأكسشن: | edsarx.2209.12561 |
قاعدة البيانات: | arXiv |
الوصف غير متاح. |