ERN-Net通过动态推理节点提升文档二值化效果，尤其擅长处理模糊笔画和噪声背景。
ERN-Net : Evolving Reason Node-Net for Document Binarization
- 采用动态演化推理节点与多尺度推理机制增强退化区域
- 在低数据低内存下性能优于现有方法，ConvNeXt-Tiny为最佳模型选择
- 基于DIBCO预训练可提效且不增加内存，仅需额外1.5小时训练
本文提出ERN-Net，一种用于高效文档图像二值化的进化推理节点网络。该方法通过动态演化推理节点与多尺度推理机制，显著增强模糊笔画、断裂字符及噪声背景等退化敏感区域的处理能力。我们对比了ResNet-101、ConvNeXt-Tiny与ConvNeXt-Base，发现ConvNeXt-Tiny在准确率与内存消耗间提供了最佳实际平衡。此外，基于DIBCO的预训练能有效提升二值化性能，且无需增加模型内存，仅需约1.5小时额外训练时间。在符合DIBCO风格的基准测试中，ERN-Net在低数据与低内存条件下均表现出色。
原文摘要 · Abstract (English)
This paper presents ERN-Net, an Evolving Reason Node-Net for efficient document image binarization. ERN-Net enhances degradation-sensitive regions, such as faint strokes, broken characters, and noisy backgrounds, through evolving reason nodes and multi-scale reasoning. We further compare ResNet-101, ConvNeXt-Tiny, and ConvNeXt-Base, and find that ConvNeXt-Tiny provides the best practical trade-off between accuracy and memory usage. In addition, DIBCO-based pretraining improves binarization performance without increasing model memory consumption, requiring only about 1.5 additional training hours. Experiments on DIBCO-style benchmarks show that ERN-Net is effective under low-data and low-memory settings.
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。