Policy Indications: How Materials Science Helps Contain Contain Covid-19 Spread  |  National Edu News: IIT Hyderabad and PharmCADD signed a pact for the co-development of new drugs   |  Teacher Insights: Be Game  |  Health Monitor: Understanding ‘Haemorrhage'  |  National Edu News: Pallikkutam GlobalConnect#3 on 'Innovative Tools for Effective Teaching'  |  Expert Counsel: The Nine Dash Line  |  National Edu News: Astronomers Find One Group of Appearing and Disappearing Stars  |  Teacher Insights: Bird Book for Children to Love Nature  |  International Edu News: New Model to Fight Social Media Deep Fakes  |  Teacher Insights: Universal Lunch Makes Students Healthier  |  Teacher Insights: Physical Activity Boosts Self Regulation  |  Parent Interventions: Anti-Inflammatory Foods Reduce Blood Fats  |  Parent Interventions: New Technique to Treat ADHD  |  Parent Interventions: Reduce Lab Tests in NICU Patients  |  Parent Interventions: Switch Off  |  
July 19, 2021 Monday 12:26:56 PM IST

New Model to Fight Social Media Deep Fakes

International Edu News

Experts of Michigan State University (MSU) have partnered with Facebook on using a new reverse-engineering research method to detect and attribute deepfakes. Technological advancements make it nearly impossible to tell whether an image of a person that appears on social media platforms is actually a real human. The new framework uses fingerprint estimation to predict network architecture and loss functions of an unknown generative model given a single generated image. With model parsing, it is possible to estimate properties of the generative models used to create each deepfake, and even associate multiple deepfakes to the model that possibly produced them. Thus it provides information about each deepfake, even ones where no prior information existed. Current methods focus on distinguishing a real image versus deepfake image relies on pre-existing knowledge while the new approach has put together a fake image data set with 100,000 synthetic images generated from 100 publicly available generative models.

Comments