COS 429 - Computer Vision

Fall 2019

Assignment 4: Deep Learning

Changelog and Clarifications

• 11/24/2019: Add fully_connected_backprop_gt.pyc into a4_part2_starter.zip; rename question 13 to question 12 in the part3 notebook.
• 11/25/2019: Update submitting requirement: README.pdf should contain both written portions and the modified code snippets.

1. Implementing SGD (10 Points)
2. Implementing Backpropogation (40 Points)
3. Learning CNNs with PyTorch (34 Points)
4. Classifying images (16 Points)

Submitting:

This assignment is due Thursday, December 5, 2019 at 11:59 PM. Please see the general notes on submitting your assignments, as well as the late policy and the collaboration policy.

As stated, our submissions this year will be done through Gradescope. The code to register for our class and how to submit can be found in the "general notes on submitting..." link above. This assignment has 2 submissions:

1. Assignment 4 Written: Submit one single README.PDF containing all written portions and your modified / added code of the assignments. Everything including both the written part and the modified code should be in this PDF file.
2. Assignment 4 Code: Submit all your code as required in each part.

Note that:

1. Assignment 4 Written is worth 100 points and Assignment 4 Code is worth 0 points on gradescope. This does not indicate we do not grade the coding part; we just put all the gradings in the written part on gradescope.
2. For all the questions that require submitting your code, please also make a screenshot or copy and poste your added / modified code snippets into your README.pdf.