YOLOv8n-GBE: A Hybrid YOLOv8n Model With Ghost Convolutions and BiFPN-ECA Attention for Solar PV Defect Localization

Reliable photovoltaic (PV) module defect detection is essential for maintaining long term energy efficiency and lowering solar power system maintenance costs. The deep learning model presented in this research is based on a hybridized YOLOv8n architecture and is lightweight and high performing. It i...

Full description

Saved in:
Bibliographic Details
Main Authors: Likitha Reddy Yeddula, Archana Pallakonda, Rayappa David Amar Raj, Rama Muni Reddy Yanamala, K. Krishna Prakasha, Mallempati Sunil Kumar
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11059241/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849415317797732352
author Likitha Reddy Yeddula
Archana Pallakonda
Rayappa David Amar Raj
Rama Muni Reddy Yanamala
K. Krishna Prakasha
Mallempati Sunil Kumar
author_facet Likitha Reddy Yeddula
Archana Pallakonda
Rayappa David Amar Raj
Rama Muni Reddy Yanamala
K. Krishna Prakasha
Mallempati Sunil Kumar
author_sort Likitha Reddy Yeddula
collection DOAJ
description Reliable photovoltaic (PV) module defect detection is essential for maintaining long term energy efficiency and lowering solar power system maintenance costs. The deep learning model presented in this research is based on a hybridized YOLOv8n architecture and is lightweight and high performing. It is designed for multi scale defect identification in a variety of imaging modalities, such as RGB, grayscale, and infrared datasets. The proposed approach combines a BiFPN based neck, Ghost Bottlenecks, and Efficient Channel Attention (ECA) to improve multi scale representation, decrease redundant computation, and increase feature extraction. The model performs better in terms of detection accuracy and efficiency, as shown by experimental findings on three benchmark datasets: PVEL-AD, PV-Multi-Defect, Solar Panel Anomalies. The model&#x2019;s respective <inline-formula> <tex-math notation="LaTeX">$\text {mAP@50}$ </tex-math></inline-formula> values are 96.5%, 94.6%, and 97.6%. At a steady inference time of only 1.9 ms and 8.1 GFLOPs, it also achieves near-perfect recall (up to 99.0%) and high precision (up to 98.4%). With just 3M parameters, the proposed hybrid model provides a much better accuracy-latency trade-off 61 current state-of-the-art models, which makes it perfect for real-time solar inspection applications, such as edge deployment in drones and embedded systems. The outcomes confirm that reliable PV fault localization under a range of operating situations may be achieved by combining deep feature fusion, lightweight attention, and efficient convolution.
format Article
id doaj-art-6efbd81f6ed043d490f6b5c58537dcdf
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-6efbd81f6ed043d490f6b5c58537dcdf2025-08-20T03:33:34ZengIEEEIEEE Access2169-35362025-01-011311401211402810.1109/ACCESS.2025.358424911059241YOLOv8n-GBE: A Hybrid YOLOv8n Model With Ghost Convolutions and BiFPN-ECA Attention for Solar PV Defect LocalizationLikitha Reddy Yeddula0Archana Pallakonda1https://orcid.org/0009-0005-7865-4156Rayappa David Amar Raj2https://orcid.org/0000-0002-5888-5513Rama Muni Reddy Yanamala3https://orcid.org/0009-0007-9132-4914K. Krishna Prakasha4https://orcid.org/0000-0001-7797-1399Mallempati Sunil Kumar5Amrita School of Artificial Intelligence, Amrita Vishwa Vidyapeetham, Coimbatore, Tamil Nadu, IndiaDepartment of Computer Science and Engineering, National Institute of Technology Warangal, Warangal, Telangana, IndiaAmrita School of Artificial Intelligence, Amrita Vishwa Vidyapeetham, Coimbatore, Tamil Nadu, IndiaDepartment of Electronics and Communication Engineering, Indian Institute of Information Technology Design and Manufacturing (IIITD&#x0026;M), Kancheepuram, Chennai, IndiaSchool of Computer Engineering, Manipal Institute of Technology (MIT), Manipal Academy of Higher Education (MAHE), Manipal, Karnataka, IndiaDepartment of Electrical and Electronics Engineering, DVR &#x0026; Dr. HS MIC College of Technology, Vijayawada, Andhra Pradesh, IndiaReliable photovoltaic (PV) module defect detection is essential for maintaining long term energy efficiency and lowering solar power system maintenance costs. The deep learning model presented in this research is based on a hybridized YOLOv8n architecture and is lightweight and high performing. It is designed for multi scale defect identification in a variety of imaging modalities, such as RGB, grayscale, and infrared datasets. The proposed approach combines a BiFPN based neck, Ghost Bottlenecks, and Efficient Channel Attention (ECA) to improve multi scale representation, decrease redundant computation, and increase feature extraction. The model performs better in terms of detection accuracy and efficiency, as shown by experimental findings on three benchmark datasets: PVEL-AD, PV-Multi-Defect, Solar Panel Anomalies. The model&#x2019;s respective <inline-formula> <tex-math notation="LaTeX">$\text {mAP@50}$ </tex-math></inline-formula> values are 96.5%, 94.6%, and 97.6%. At a steady inference time of only 1.9 ms and 8.1 GFLOPs, it also achieves near-perfect recall (up to 99.0%) and high precision (up to 98.4%). With just 3M parameters, the proposed hybrid model provides a much better accuracy-latency trade-off 61 current state-of-the-art models, which makes it perfect for real-time solar inspection applications, such as edge deployment in drones and embedded systems. The outcomes confirm that reliable PV fault localization under a range of operating situations may be achieved by combining deep feature fusion, lightweight attention, and efficient convolution.https://ieeexplore.ieee.org/document/11059241/BiFPNdefect detectionefficient channel attentionghost bottlenecksYOLOv8n
spellingShingle Likitha Reddy Yeddula
Archana Pallakonda
Rayappa David Amar Raj
Rama Muni Reddy Yanamala
K. Krishna Prakasha
Mallempati Sunil Kumar
YOLOv8n-GBE: A Hybrid YOLOv8n Model With Ghost Convolutions and BiFPN-ECA Attention for Solar PV Defect Localization
IEEE Access
BiFPN
defect detection
efficient channel attention
ghost bottlenecks
YOLOv8n
title YOLOv8n-GBE: A Hybrid YOLOv8n Model With Ghost Convolutions and BiFPN-ECA Attention for Solar PV Defect Localization
title_full YOLOv8n-GBE: A Hybrid YOLOv8n Model With Ghost Convolutions and BiFPN-ECA Attention for Solar PV Defect Localization
title_fullStr YOLOv8n-GBE: A Hybrid YOLOv8n Model With Ghost Convolutions and BiFPN-ECA Attention for Solar PV Defect Localization
title_full_unstemmed YOLOv8n-GBE: A Hybrid YOLOv8n Model With Ghost Convolutions and BiFPN-ECA Attention for Solar PV Defect Localization
title_short YOLOv8n-GBE: A Hybrid YOLOv8n Model With Ghost Convolutions and BiFPN-ECA Attention for Solar PV Defect Localization
title_sort yolov8n gbe a hybrid yolov8n model with ghost convolutions and bifpn eca attention for solar pv defect localization
topic BiFPN
defect detection
efficient channel attention
ghost bottlenecks
YOLOv8n
url https://ieeexplore.ieee.org/document/11059241/
work_keys_str_mv AT likithareddyyeddula yolov8ngbeahybridyolov8nmodelwithghostconvolutionsandbifpnecaattentionforsolarpvdefectlocalization
AT archanapallakonda yolov8ngbeahybridyolov8nmodelwithghostconvolutionsandbifpnecaattentionforsolarpvdefectlocalization
AT rayappadavidamarraj yolov8ngbeahybridyolov8nmodelwithghostconvolutionsandbifpnecaattentionforsolarpvdefectlocalization
AT ramamunireddyyanamala yolov8ngbeahybridyolov8nmodelwithghostconvolutionsandbifpnecaattentionforsolarpvdefectlocalization
AT kkrishnaprakasha yolov8ngbeahybridyolov8nmodelwithghostconvolutionsandbifpnecaattentionforsolarpvdefectlocalization
AT mallempatisunilkumar yolov8ngbeahybridyolov8nmodelwithghostconvolutionsandbifpnecaattentionforsolarpvdefectlocalization