Balancing Complexity and Performance of Machine Learning Models for Avian Pests Sound Detection in Agricultural Environments
Agricultural pest control traditionally relies on inefficient visual inspections. Acoustic monitoring offers a promising alternative by analyzing pest-specific sounds. While effective, implementing acoustic monitoring in agricultural settings faces practical constraints, particularly the limited com...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11039837/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849683281704910848 |
|---|---|
| author | Micheline Kazeneza Anna Sergeevna Bosman Destiny Kwabla Amenyedzi Damien Hanyurwimfura Emmanuel Ndashimye Anthony Vodacek |
| author_facet | Micheline Kazeneza Anna Sergeevna Bosman Destiny Kwabla Amenyedzi Damien Hanyurwimfura Emmanuel Ndashimye Anthony Vodacek |
| author_sort | Micheline Kazeneza |
| collection | DOAJ |
| description | Agricultural pest control traditionally relies on inefficient visual inspections. Acoustic monitoring offers a promising alternative by analyzing pest-specific sounds. While effective, implementing acoustic monitoring in agricultural settings faces practical constraints, particularly the limited computational resources available in remote farming environments. This necessitates optimized machine learning (ML) solutions for low-power edge devices. This study evaluates ML models for bird pest detection on resource-constrained platforms. We evaluated convolutional neural networks (CNNs), recurrent neural networks (RNNs), and traditional ML models by comparing standalone and knowledge-distilled versions of EfficientNetB0 and gated recurrent unity (GRU) against EfficientNetB4, Long short-term memory (LSTM), MobileNetV2, LightGBM, and support vector machine (SVM). Analysis revealed significant performance variations across computational requirements. LightGBM achieved 98% accuracy with minimal resources (8,500 parameters, 7KB, 0.6ms inference), demonstrating good efficiency. SVM (97% accuracy) and distilled GRU (86% accuracy) also showed favorable performance-to-resource ratios. Knowledge distillation substantially enhanced the accuracy of EfficientNetB0 (from 73% to 98%) and modestly improved GRU (from 84% to 86%). We examined platform compatibility across computing tiers, discovering that while high-performance edge devices (Jetson Nano, Raspberry Pi 4) support all studied models effectively, microcontrollers require specialized approaches. Advanced microcontrollers (such as ESP32-S3 and STM32H7) can accommodate optimized implementations, while highly constrained platforms (such as Arduino Nano) require TinyML techniques. This research contributes 1) an on-farm audio dataset, 2) comprehensive cross-model evaluation metrics, and 3) deployment optimization strategies for acoustic pest detection systems in resource-constrained agricultural environments. |
| format | Article |
| id | doaj-art-3e971b5e9ff4431cb122e1b35a71be68 |
| institution | DOAJ |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-3e971b5e9ff4431cb122e1b35a71be682025-08-20T03:23:57ZengIEEEIEEE Access2169-35362025-01-011310581310582710.1109/ACCESS.2025.358062011039837Balancing Complexity and Performance of Machine Learning Models for Avian Pests Sound Detection in Agricultural EnvironmentsMicheline Kazeneza0https://orcid.org/0009-0002-2283-9184Anna Sergeevna Bosman1https://orcid.org/0000-0003-3546-1467Destiny Kwabla Amenyedzi2https://orcid.org/0000-0002-3703-7312Damien Hanyurwimfura3https://orcid.org/0000-0002-6290-1256Emmanuel Ndashimye4https://orcid.org/0000-0003-3549-8155Anthony Vodacek5https://orcid.org/0000-0001-9196-0928African Centre of Excellence in Internet of Things, University of Rwanda, Kigali, RwandaDepartment of Computer Science, University of Pretoria, Pretoria, South AfricaAfrican Centre of Excellence in Internet of Things, University of Rwanda, Kigali, RwandaAfrican Centre of Excellence in Internet of Things, University of Rwanda, Kigali, RwandaSchool of ECE, Carnegie Mellon University Africa, Kigali, RwandaChester F. Carlson Center for Imaging Science, Rochester Institute of Technology, Rochester, NY, USAAgricultural pest control traditionally relies on inefficient visual inspections. Acoustic monitoring offers a promising alternative by analyzing pest-specific sounds. While effective, implementing acoustic monitoring in agricultural settings faces practical constraints, particularly the limited computational resources available in remote farming environments. This necessitates optimized machine learning (ML) solutions for low-power edge devices. This study evaluates ML models for bird pest detection on resource-constrained platforms. We evaluated convolutional neural networks (CNNs), recurrent neural networks (RNNs), and traditional ML models by comparing standalone and knowledge-distilled versions of EfficientNetB0 and gated recurrent unity (GRU) against EfficientNetB4, Long short-term memory (LSTM), MobileNetV2, LightGBM, and support vector machine (SVM). Analysis revealed significant performance variations across computational requirements. LightGBM achieved 98% accuracy with minimal resources (8,500 parameters, 7KB, 0.6ms inference), demonstrating good efficiency. SVM (97% accuracy) and distilled GRU (86% accuracy) also showed favorable performance-to-resource ratios. Knowledge distillation substantially enhanced the accuracy of EfficientNetB0 (from 73% to 98%) and modestly improved GRU (from 84% to 86%). We examined platform compatibility across computing tiers, discovering that while high-performance edge devices (Jetson Nano, Raspberry Pi 4) support all studied models effectively, microcontrollers require specialized approaches. Advanced microcontrollers (such as ESP32-S3 and STM32H7) can accommodate optimized implementations, while highly constrained platforms (such as Arduino Nano) require TinyML techniques. This research contributes 1) an on-farm audio dataset, 2) comprehensive cross-model evaluation metrics, and 3) deployment optimization strategies for acoustic pest detection systems in resource-constrained agricultural environments.https://ieeexplore.ieee.org/document/11039837/Acoustic monitoringbird pest detectiondeep learningknowledge distillationmachine learning |
| spellingShingle | Micheline Kazeneza Anna Sergeevna Bosman Destiny Kwabla Amenyedzi Damien Hanyurwimfura Emmanuel Ndashimye Anthony Vodacek Balancing Complexity and Performance of Machine Learning Models for Avian Pests Sound Detection in Agricultural Environments IEEE Access Acoustic monitoring bird pest detection deep learning knowledge distillation machine learning |
| title | Balancing Complexity and Performance of Machine Learning Models for Avian Pests Sound Detection in Agricultural Environments |
| title_full | Balancing Complexity and Performance of Machine Learning Models for Avian Pests Sound Detection in Agricultural Environments |
| title_fullStr | Balancing Complexity and Performance of Machine Learning Models for Avian Pests Sound Detection in Agricultural Environments |
| title_full_unstemmed | Balancing Complexity and Performance of Machine Learning Models for Avian Pests Sound Detection in Agricultural Environments |
| title_short | Balancing Complexity and Performance of Machine Learning Models for Avian Pests Sound Detection in Agricultural Environments |
| title_sort | balancing complexity and performance of machine learning models for avian pests sound detection in agricultural environments |
| topic | Acoustic monitoring bird pest detection deep learning knowledge distillation machine learning |
| url | https://ieeexplore.ieee.org/document/11039837/ |
| work_keys_str_mv | AT michelinekazeneza balancingcomplexityandperformanceofmachinelearningmodelsforavianpestssounddetectioninagriculturalenvironments AT annasergeevnabosman balancingcomplexityandperformanceofmachinelearningmodelsforavianpestssounddetectioninagriculturalenvironments AT destinykwablaamenyedzi balancingcomplexityandperformanceofmachinelearningmodelsforavianpestssounddetectioninagriculturalenvironments AT damienhanyurwimfura balancingcomplexityandperformanceofmachinelearningmodelsforavianpestssounddetectioninagriculturalenvironments AT emmanuelndashimye balancingcomplexityandperformanceofmachinelearningmodelsforavianpestssounddetectioninagriculturalenvironments AT anthonyvodacek balancingcomplexityandperformanceofmachinelearningmodelsforavianpestssounddetectioninagriculturalenvironments |