Multimodal Collaboration via Multi-stage Augmentation for Few-Shot Foodborne Pathogens Classification.

Saved in:

Bibliographic Details
Title:	Multimodal Collaboration via Multi-stage Augmentation for Few-Shot Foodborne Pathogens Classification.
Authors:	Zhang, Yixiang¹ 1545032733@qq.com, Hang, Yuting¹ 2763626778@qq.com, Fang, Jie² fangjie@xupt.edu.cn
Source:	IAENG International Journal of Computer Science. Jul2026, Vol. 53 Issue 7, p2825-2833. 9p.
Subjects:	Food pathogens, Data augmentation, Public health, Feature extraction, Machine learning
Abstract:	Rapid detection of foodborne pathogens is critical for public health. While machine learning enables efficient pathogen detection, data scarcity and annotation challenges degrade few-shot classification performance. Traditional few-shot methods (e.g., meta-learning) struggle due to insufficient prior knowledge. To address this, we propose a multimodal collaboration framework with three components: (1) Multimodal Data Augmentation (MDA) using scale variation, noise, and image mixing to alleviate sample scarcity. (2) Multimodal Feature Representation (MFR) fusing shallow contour and deep semantic features via multi-scale concatenation and channel attention. (3) Multimodal Collaborative Decision (MCD) employing majority voting to integrate predictions and reduce interferences. Experiments show our method achieves 74.46% and 80.77% accuracy in 3-shot and 5-shot tasks, respectively, outperforming state-of-the-art approaches. This work provides an effective solution for few-shot pathogen detection, enhancing food safety supervision. [ABSTRACT FROM AUTHOR]
	Copyright of IAENG International Journal of Computer Science is the property of International Association of Engineers (IAENG) and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	Rapid detection of foodborne pathogens is critical for public health. While machine learning enables efficient pathogen detection, data scarcity and annotation challenges degrade few-shot classification performance. Traditional few-shot methods (e.g., meta-learning) struggle due to insufficient prior knowledge. To address this, we propose a multimodal collaboration framework with three components: (1) Multimodal Data Augmentation (MDA) using scale variation, noise, and image mixing to alleviate sample scarcity. (2) Multimodal Feature Representation (MFR) fusing shallow contour and deep semantic features via multi-scale concatenation and channel attention. (3) Multimodal Collaborative Decision (MCD) employing majority voting to integrate predictions and reduce interferences. Experiments show our method achieves 74.46% and 80.77% accuracy in 3-shot and 5-shot tasks, respectively, outperforming state-of-the-art approaches. This work provides an effective solution for few-shot pathogen detection, enhancing food safety supervision. [ABSTRACT FROM AUTHOR]
ISSN:	1819656X