Confidence-Guided Diffusion Augmentation for Enhanced Bangla Compound Character Recognition

TL;DR

This paper introduces a confidence-guided diffusion augmentation method that synthesizes high-quality handwritten Bangla compound characters, significantly improving recognition accuracy in low-resource scenarios.

Contribution

It proposes a novel diffusion-based augmentation framework with classifier guidance and quality filtering, enhancing recognition of complex Bangla compound characters.

Findings

Achieved 89.2% accuracy on AIBangla dataset, surpassing previous benchmarks.

Demonstrated consistent improvements across multiple neural network architectures.

Validated effectiveness of quality-aware synthetic data in low-resource script recognition.

Abstract

Recognition of handwritten Bangla compound characters remains a challenging problem due to complex character structures, large intra-class variation, and limited availability of high-quality annotated data. Existing Bangla handwritten character recognition systems often struggle to generalize across diverse writing styles, particularly for compound characters containing intricate ligatures and diacritical variations. In this work, we propose a confidence-guided diffusion augmentation framework for low-resolution Bangla compound character recognition. Our framework combines class-conditional diffusion modeling with classifier guidance to synthesize high-quality handwritten compound character samples. To further improve generation quality, we introduce Squeeze-and-Excitation enhanced residual blocks within the diffusion model's U-Net backbone. We additionally propose a confidence-based filtering mechanism where pre-trained classifiers act as quality gates to retain only highly class-consistent synthetic samples. The filtered synthetic images are fused with the original training data and used to retrain multiple classification architectures. Experiments conducted on the AIBangla compound character dataset demonstrate consistent performance improvements across ResNet50, DenseNet121, VGG16, and Vision Transformer architectures. Our best-performing model achieves 89.2\% classification accuracy, surpassing the previously published AIBangla benchmark by a substantial margin. The results demonstrate that quality-aware diffusion augmentation can effectively enhance handwritten character recognition performance in low-resource script domains.