This paper addresses the challenge of accurately detecting defects in miniature capacitors, which are crucial for high-precision charge storage and voltage regulation in electronic devices. Traditional methods, such as microscopic inspection and electrical performance tests, are limited by their inability to detect defects at the micron scale and their reliance on skilled operators. To overcome these limitations, the authors propose an enhanced version of the 'you only look once' (YOLOv8) architecture, specifically tailored for micro-capacitor defect inspection. The key contributions of this work include: 1. **Enhanced YOLOv8 Architecture**: The authors integrate the bidirectional feature pyramid network (BiFPN) and the simplified attention module (SimAM) into the YOLOv8 model. BiFPN improves multi-scale feature fusion, while SimAM enhances the model's ability to focus on important features, particularly small defects. 2. **WISE-IOU Loss Function**: The model uses the weighted intersection over union (WISE-IOU) loss function, which replaces the traditional CIoU loss function. This modification reduces the negative impact of low-quality samples during training, improving the model's generalization and robustness. 3. **Dataset and Evaluation**: A high-quality micro-capacitor surface defect (MCSD) dataset was constructed, containing 1358 images of four distinct defect types. The model achieved a mean average precision (mAP) of 95.8% at a threshold of 0.5, representing a 9.5% improvement over the original YOLOv8 architecture. 4. **Performance Analysis**: The experimental results show that the proposed model outperforms other state-of-the-art algorithms like SSD, Faster R-CNN, and YOLOv5n, v7-tiny, and v8. The model maintains a real-time detection speed suitable for industrial applications, with a frame rate of 78 FPS. 5. **Conclusion**: The enhanced YOLOv8 model demonstrates significant progress in defect detection for miniature capacitors, offering a robust and efficient solution for quality control in electronic manufacturing. Future work will focus on expanding the dataset and fine-tuning the system to handle a broader range of defects and industrial environments.This paper addresses the challenge of accurately detecting defects in miniature capacitors, which are crucial for high-precision charge storage and voltage regulation in electronic devices. Traditional methods, such as microscopic inspection and electrical performance tests, are limited by their inability to detect defects at the micron scale and their reliance on skilled operators. To overcome these limitations, the authors propose an enhanced version of the 'you only look once' (YOLOv8) architecture, specifically tailored for micro-capacitor defect inspection. The key contributions of this work include: 1. **Enhanced YOLOv8 Architecture**: The authors integrate the bidirectional feature pyramid network (BiFPN) and the simplified attention module (SimAM) into the YOLOv8 model. BiFPN improves multi-scale feature fusion, while SimAM enhances the model's ability to focus on important features, particularly small defects. 2. **WISE-IOU Loss Function**: The model uses the weighted intersection over union (WISE-IOU) loss function, which replaces the traditional CIoU loss function. This modification reduces the negative impact of low-quality samples during training, improving the model's generalization and robustness. 3. **Dataset and Evaluation**: A high-quality micro-capacitor surface defect (MCSD) dataset was constructed, containing 1358 images of four distinct defect types. The model achieved a mean average precision (mAP) of 95.8% at a threshold of 0.5, representing a 9.5% improvement over the original YOLOv8 architecture. 4. **Performance Analysis**: The experimental results show that the proposed model outperforms other state-of-the-art algorithms like SSD, Faster R-CNN, and YOLOv5n, v7-tiny, and v8. The model maintains a real-time detection speed suitable for industrial applications, with a frame rate of 78 FPS. 5. **Conclusion**: The enhanced YOLOv8 model demonstrates significant progress in defect detection for miniature capacitors, offering a robust and efficient solution for quality control in electronic manufacturing. Future work will focus on expanding the dataset and fine-tuning the system to handle a broader range of defects and industrial environments.