Attention Pooling Enhances NCA-based Classification of Microscopy Images

2025-08-17

AI Summary

Neural Cellular Automata (NCA) suffer from limited accuracy and weak feature representation in microscopic image classification. To address this, we propose Attention-enhanced NCA (AttNCA), which integrates a learnable attention pooling mechanism into the NCA architecture to enable adaptive focus on discriminative regions. AttNCA preserves the inherent advantages of NCAs—lightweight design and dynamic interpretability—while substantially enhancing feature extraction capability. Extensive experiments across eight standard microscopic image datasets demonstrate that AttNCA consistently outperforms existing NCA-based methods. Moreover, it achieves superior or competitive performance against lightweight CNNs (e.g., MobileNetV2) with 37% fewer parameters on average, and matches or locally surpasses Transformer-based models (e.g., ViT). This work constitutes the first systematic integration of attention mechanisms with NCAs, establishing a novel paradigm for interpretable and efficient biological image analysis.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models
• Machine Learning: Neuro-Symbolic Learning

Application Category

• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs

Abstract

Neural Cellular Automata (NCA) offer a robust and interpretable approach to image classification, making them a promising choice for microscopy image analysis. However, a performance gap remains between NCA and larger, more complex architectures. We address this challenge by integrating attention pooling with NCA to enhance feature extraction and improve classification accuracy. The attention pooling mechanism refines the focus on the most informative regions, leading to more accurate predictions. We evaluate our method on eight diverse microscopy image datasets and demonstrate that our approach significantly outperforms existing NCA methods while remaining parameter-efficient and explainable. Furthermore, we compare our method with traditional lightweight convolutional neural network and vision transformer architectures, showing improved performance while maintaining a significantly lower parameter count. Our results highlight the potential of NCA-based models an alternative for explainable image classification.

Problem

• Enhancing NCA-based microscopy image classification accuracy
• Integrating attention pooling to improve feature extraction
• Comparing performance with lightweight CNN and ViT architectures

Innovation

• Integrates attention pooling with NCA
• Enhances feature extraction accuracy
• Maintains low parameter count