Week 7 High-Level Vision Flashcards
What does high-level vision involve?
High-level vision involves understanding scenes, objects, and relationships in images.
How does high-level vision build on lower levels?
It builds on low-level and mid-level vision tasks like edge detection and segmentation.
What is scene understanding?
Scene understanding predicts the overall context or layout of an image.
What does object recognition do?
Object recognition identifies objects and assigns labels based on learned features.
What is the purpose of object detection?
Object detection locates objects and predicts their bounding boxes.
How does semantic segmentation classify pixels?
Semantic segmentation assigns a label to every pixel, identifying classes.
What is instance segmentation?
Instance segmentation separates individual objects within the same class.
What does relationship detection predict?
Relationship detection predicts how objects interact or relate within a scene.
What is image captioning?
Image captioning generates textual descriptions of an image’s content.
What is the application of face recognition?
Face recognition is a specific high-level task for identifying individuals in images.
