OCTOBER MEETING: Tuesday, October 29, 2024 at 5:20pm (US Central)

LOCATION:
In person: Chambers Hall, Lower Level
600 Foster Steet, Evanston Campus

AGENDA:
5:20pm - Meet and Greet
5:30pm - Talk by Matt Groh, Assistant Professor, Kellogg School of Management
“Deepfake Detection by Humans Across Contexts Ranging from Political Speeches to Human Photography to Podcasts”
6:30pm - Q&A

SPEAKER:
Matt Groh, Assistant Professor, Kellogg School of Management; Core Faculty, Northwestern Institute on Complex Systems; and Principal Investigator, Human-AI Collaboration Lab

ABSTRACT:
Deepfake Detection by Humans across Contexts Ranging from Political Speeches to Human Photography to Podcasts

This talk will examine people’s ability to detect deepfakes and other AI-generated content across several contexts. We will discuss the latest advances in perceptually realistic synthetic media, share insights from a paper under review on evaluating photorealism in diffusion models, and deep dive into an experiment examining human detection of political speech deepfakes. We’ll present evidence that audio and visual information enables more accurate discernment than text alone, which means human discernment relies more on how something is said, the audio-visual cues, than what is said, the speech content. We’ll also show how deepfakes with audio produced by the state-of-the-art text-to-speech algorithms are harder to discern than the same deepfakes with voice actor audio.

DATA SCIENCE NIGHTS are monthly talks on data science techniques or applications, organized by Northwestern University graduate students and scholars. Aspiring, beginning, and advanced data scientists are welcome! For more information: http://bit.ly/nico-dsn

Related Info