Search papers, labs, and topics across Lattice.
This paper introduces ERN-Net, a novel Evolving Reason Node-Net designed for efficient document image binarization, specifically targeting degradation-sensitive areas such as faint strokes and noisy backgrounds. By leveraging evolving reason nodes and multi-scale reasoning, the model enhances binarization performance while maintaining low memory usage. Comparative experiments reveal that ConvNeXt-Tiny strikes the optimal balance between accuracy and memory efficiency, with DIBCO-based pretraining further boosting performance without significant resource overhead.
ERN-Net achieves superior document binarization performance in low-data and low-memory scenarios, making it a game-changer for resource-constrained applications.
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.
