final project is to build a machine learning system capable of predicting or filling in these pixels

INSTRUCTIONS TO CANDIDATES

ANSWER ALL QUESTIONS

Figure 1: Two images, 900 x 900 pixels, with the middle 1/9th removed.

In each of these images, the middle 300x300 pixels have been removed (rather, have been painted white). The masked files are available with this PDF. The goal of this final project is to build a machine learning system capable of predicting or filling in these pixels, to 'realistically' complete the image. There are many ways this might be approached, for example trying to find a relationship between related or nearby pixels, and using that to fill in missing pixels.

(surrounding visual context data) → (missing visual data)

(1)

In building a system to accomplish this, you need to consider the following elements:

• Input Space: What are you taking as input? How is it formatted and structured? Notice that form image data, spatial relationships matter. Is your input numerical or categorial? What training data do you have available to use?

⚫ Output Space: What are you taking as output? Do you want to predict one pixel at a time, or multiple pixels, or the entire patch? Does your output need to be numerical, or categorial? What training data do you have available to use?

⚫ Model Space: For your choice of input space and output space, you need to consider a model that maps between them. What kind of models make sense? What kind of models are easier or harder to work with? Will your model be expressible enough to capture the relationship between input and output? How do you know if your model is too expressible?

⚫ Loss: What makes one model better than another? Notice that in this problem the quality of a model is perceptual / aesthetic - how nice does the result look? But we need a numerical measure of loss to make any of our tools work.

• Training Algorithm: For your choice of loss function, you need an algorithm to produce a good model. What is applicable to your choice of model, loss, and data? What are potential problems to consider, or choices you have to make for your algorithm?

 

Attachments:

Related Questions

. Introgramming & Unix Fall 2018, CRN 44882, Oakland University Homework Assignment 6 - Using Arrays and Functions in C

DescriptionIn this final assignment, the students will demonstrate their ability to apply two ma

. The standard path finding involves finding the (shortest) path from an origin to a destination, typically on a map. This is an

Path finding involves finding a path from A to B. Typically we want the path to have certain properties,such as being the shortest or to avoid going t

. Develop a program to emulate a purchase transaction at a retail store. This program will have two classes, a LineItem class and a Transaction class. The LineItem class will represent an individual

Develop a program to emulate a purchase transaction at a retail store. Thisprogram will have two classes, a LineItem class and a Transaction class. Th

. SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of Sea Ports. Here are the classes and their instance variables we wish to define:

1 Project 1 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of

. Project 2 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of Sea Ports. Here are the classes and their instance variables we wish to define:

1 Project 2 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of