Microsoft Certified Azure AI Engineer Associate - AI-102
Get started today
Ultimate access to all questions.
You have trained a custom object detection model to detect defects in manufactured parts. The model has been trained on a diverse set of images but is still underperforming in detecting subtle defects. What strategies would you employ to enhance the model's ability to detect these subtle defects, including potential modifications to the training data and model architecture?
You have trained a custom object detection model to detect defects in manufactured parts. The model has been trained on a diverse set of images but is still underperforming in detecting subtle defects. What strategies would you employ to enhance the model's ability to detect these subtle defects, including potential modifications to the training data and model architecture?
Simulated
Explanation:
To detect subtle defects, increasing the complexity of the model architecture can help the model learn more intricate patterns. Adding more diverse images and applying data augmentation techniques can enhance the model's ability to generalize from the data, making it more effective at detecting subtle differences.