Give Your Pi Eyes: Machine Learning with Image Models
Spring 2021 | HCDE 496/596 | 2 credits
Meetings: Wednesday, 4:00 – 5:30 pm, in Zoom
Adi Azulay, HCDE Affiliate Instructor | Office of the CTO, Microsoft Corporation
Andrew Davidson, HCDE Associate Teaching Professor
Machine Learning (ML) is all around us and impacts our lives in many ways, from email spam filters to facial recognition to fraud detection on credit cards. Until recently training, and running ML models required a team of data scientists and mathematicians. Now with software like Lobe and Custom Vision all you need to do is upload images and labels and the software does the rest.
In this DRG students will train their own ML models using Lobe and deploy them on a Raspberry Pi (hardware will be provided). As a group we’ll examine how ML models are built, what type of data goes into them, and dig into the ethics and efficacy of using them in different applications.
We are looking for dedicated and enthusiastic students at all levels (BS to PhD) to work on this project. Technical skill is not required, and students with a background in humanities or liberal arts are encouraged to apply.
Applicants should be familiar with basic concepts of programming (as covered in HCDE 310 or HCDE 524, for instance). Experience programming in Python is desirable but not required. All applicants should have already taken HCDE 318 or HCDE 518.
The DRG will meet on Wednesdays, from 4:00 – 5:30 pm. All participants are expected to register for 2 credits.
To apply, please submit your resume and a letter explaining your interest and qualifications for the project here: application form.
The deadline for applications is March 12, 2021. We will notify accepted students by March 16. If you are accepted, we will provide you information about how you will receive the hardware kit you will use in the DRG.
For questions please contact either of us: