High-Accuracy Recognition Method for Diseased Chicken Feces Based on Image and Text Information Fusion
Related Experiment Video
Updated: Jul 13, 2026
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
Published on: April 14, 2023
Duanli Yang1,2, Zishang Tian1,2, Jianzhong Xi3
1College of Information Science and Technology, Hebei Agricultural University, Baoding 071001, China.
Abstract:
Poultry feces, a critical biomarker for health assessment, requires timely and accurate pathological identification for food safety. Conventional visual-only methods face limitations due to environmental sensitivity and high visual similarity among feces from different diseases. To address this, we propose MMCD (Multimodal Chicken-feces Diagnosis), a ResNet50-based multimodal fusion model leveraging semantic complementarity between images and descriptive text to enhance diagnostic precision. Key innovations include the following: (1) Integrating MASA(Manhattan self-attention)and DSconv (Depthwise Separable convolution) into the backbone network to mitigate feature confusion. (2) Utilizing a pre-trained BERT to extract textual semantic features, reducing annotation dependency and cost. (3) Designing a lightweight Gated Cross-Attention (GCA) module for dynamic multimodal fusion, achieving a 41% parameter reduction versus cross-modal transformers. Experiments demonstrate that MMCD significantly outperforms single-modal baselines in Accuracy (+8.69%), Recall (+8.72%), Precision (+8.67%), and F1 score (+8.72%). It surpasses simple feature concatenation by 2.51-2.82% and reduces parameters by 7.5M and computations by 1.62 GFLOPs versus the base ResNet50. This work validates multimodal fusion's efficacy in pathological fecal detection, providing a theoretical and technical foundation for agricultural health monitoring systems.
Insights
This study introduces a multimodal AI model for diagnosing poultry diseases using chicken feces images and text. The novel approach significantly improves diagnostic accuracy, offering a robust tool for food safety and agricultural health monitoring.
Area of Science:
- Agricultural Science
- Computer Science
- Veterinary Medicine
Background:
- Poultry feces analysis is vital for food safety and disease detection.
- Current visual methods for fecal analysis are limited by environmental factors and disease similarity.
- Accurate pathological identification of poultry feces is crucial for timely intervention.
Purpose of the Study:
- To develop a multimodal fusion model for enhanced pathological identification of chicken feces.
- To improve diagnostic accuracy beyond conventional single-modal approaches.
- To reduce annotation dependency and computational costs in fecal analysis.
Main Methods:
- Proposed MMCD (Multimodal Chicken-feces Diagnosis), a ResNet50-based model integrating image and text data.
- Incorporated Manhattan self-attention (MASA) and Depthwise Separable convolution (DSconv) to refine feature extraction.
- Employed a pre-trained BERT for text feature extraction and a Gated Cross-Attention (GCA) module for efficient multimodal fusion.
Main Results:
- MMCD significantly outperformed single-modal baselines in Accuracy (+8.69%), Recall (+8.72%), Precision (+8.67%), and F1 score (+8.72%).
- Achieved a 41% parameter reduction compared to standard cross-modal transformers.
- Demonstrated superior performance over simple feature concatenation by 2.51-2.82%.
Conclusions:
- Multimodal fusion is highly effective for pathological fecal detection in poultry.
- The MMCD model offers a robust, efficient, and accurate solution for agricultural health monitoring.
- This research provides a foundation for advanced AI-driven systems in food safety and animal health.
