In the image recognition problem I am trying to solve, we are supposed to develop a model that predicts the cost of a house based on images of several rooms in the house. Each house has several images but not all houses have the same number of photos. Some houses have more photos, some have less. How can one setup a CNN to use such data? Unfortunately, I cannot share any data because this is related to a problem I am solving at work. Can someone point me towards sample code/problems/kaggle competitions that deal with this problem?

Full article content could not be extracted automatically. Read the original below.