Join Mark Niemann-Ross for an in-depth discussion in this video Introduction: Facial recognition, part of Code Clinic: R.
(computer beeping) (hard drive whirring) - [Instructor] Hello, and welcome to the fifth problem of the Code Clinic series. This is a challenge to find faces in a photograph. 50 years ago this problem would have been difficult, or impossible. The technology, and algorithms for facial recognition were just being created. Computers were barely powerful enough to run the algorithms, and it could take hours to recognize one face. Today we have real-time performance, and use it as a security device on some mobile computers.
In this challenge we're asking our authors to demonstrate how to do this with their programming language. We've challenged our authors to create a function that does the following. First the function should accept a JPEG, GIF, or PNG file of varying dimensions. Their function should then identify the number of faces in that image. There may not be any faces in which case the function should return zero. If the author chooses to go for extra credit their function can create a copy of the image with the faces visually identified either with a surrounding box, or other method.
The function should return a JSON file with two fields CountFaces, and imageLocation. Before you go any further take some time, and solve the problem yourself. You'll get a better appreciation for the challenge our authors face. Get it? Face. Okay, authors, let's find out who's who.
- Identify when to use the vector model.
- Describe how to efficiently store results.
- Explain the functionality TCL/TK provides.
- Summarize dashboard solutions.
- Cite the role of the Shiny Server.
- Name the package containing Date Intervals.