Semantic Segmentation Quiz

What you'll be asked

1. What is the primary objective of semantic segmentation in computer vision?
2. When deploying a custom semantic segmentation model using Azure Machine Learning, which data labeling format is typically required for training?
3. Which architecture is most commonly associated with high-performance semantic segmentation tasks?
4. In the context of Azure computer vision, why would you choose an 'Instance Segmentation' model over a 'Semantic Segmentation' model?
5. When training a semantic segmentation model on Azure, you notice the model is performing well on training data but failing on validation data due to class imbalance (e.g., background pixels significantly outnumbering target object pixels). Which loss function modification is most appropriate?

Answers and explanations come after you submit.