Synapse
⌘+K
Synapse
PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
June 11, 2026Autonomous Intelligent SystemsOpen Access

Diffusion-augmented YOLO26-Swin cascaded framework with hybrid SHAP-CAM for autonomous power grid inspection

View Full Paper
Ask AI
Bookmark
Share

Authors

SSStefano Frizzo StefenonJMJoão P. Matos-CarvalhoVMViviana Cocco Mariani

Discussion

Loading...

Member takes

Overview

Randomized trial demonstrates enhanced fault classification in power grid inspection, indicating improved safety and reliability.

Key Points

  • The aim is to develop a streamlined framework for autonomous inspection of power grid insulators using advanced deep learning techniques.
  • Utilized a conditional diffusion model for synthetic fault image generation to address data imbalance.
  • Implemented a YOLO26-Swin architecture for robust insulator detection and fault classification.
  • Employed SHAP-CAM for visual explanations of the model's predictions.
  • Achieved an F1-score of 0.98149, indicating high model accuracy.
  • Reached a mean Average Precision (mAP)@[0.5] of 0.98951, surpassing existing models.
  • Validated the effectiveness of diffusion models for data augmentation and improved interpretability.

Cite This Study

Stefenon et al. (2026) studied this question.

synapsesocial.com/papers/6a2a505d80c8f91e7f39ce3fhttps://doi.org/10.1007/s43684-026-00135-2
View Full Paper
Ask AI
Bookmark
Share