This is a bonus episode of the podcast in a two part series covering a Nisos project with New York University (NYU) on the technical considerations for creating deepfakes as well as processes and procedures to detect them featuring Nisos Senior Software Engineer Dev Badlu.
Outline:
- (00:28) Introduction
- (01:06) Overview of Nisos Deepfake Project
- (02:21) Question 1 - What kind of research did you have to do to start making deepfake videos?
- (03:51) Question 2 - Are there a lot of open source libraries intended to make building deepfakes easier?
- (05:20) Question 3 - What was the most difficult part of the process for building a deepfake video?
- (06:15) Question 4 - How difficult is it to make deepfake audio?
- (06:56) Question 5 - How can we detect deepfakes?
