CS1674: Homework 7 - Written

1. Name three examples of recognition problems in computer vision. For each, explain in 1-2 sentences what the task that we're trying to solve in each of those problems is (i.e. elaborate on what the problem means).
2. Consider generic categorization versus instance recognition. Why might the former be more challenging than the latter? Why might the reverse be true in some cases?
3. Briefly describe what we do during recognition at training time, and what we do at test time.
4. How does a K-Nearest Neighbor classifier work?
5. Why might a Spatial Pyramid Match make sense for scene recognition but not object recognition?