Sensetime has created the "largest face forgery detection dataset" ever to find deepfakes
Sensetime Research has created a new benchmark for detecting deepfakes using high quality commissioned videos and a dataset of convincing deepfakes its researchers made themselves.
The benchmark and dataset, called DeeperForensics-1.0, contains more training videos than other detection systems, and uses footage filmed in various lighting conditions and angles.
Sensetime partnered with researchers from Nanyang Technological University in Singapore on this tool.
The videos they commissioned included 100 actors from 26 different nationalities ranging in age between 20 and 45, which they then used to create deepfakes in YouTube videos. They then distorted the manipulated footage and fed it into the dataset, giving it the training it needs to identify tampered videos.
Facebook and Microsoft began a similar initiative in 2019, when they announced that they'd commissioned videos, created deepfakes, and would challenge researchers with finding ways to use that data to find patterns common among fake videos.
AI Research and Development →
Follow the history and current research climate of AI.
Brilliant offers courses in computer science, math, and natural sciences.
Brilliant is made with the loving efforts of lifelong learners from MIT, Caltech, Duke, the University of Chicago, and more.
In school, people are often trained to apply formulas to rote problems. But this traditional approach prevents deeper understanding of concepts, reduces independent critical thinking, and cultivates few useful skills.
Whether you're looking for Computer Science Fundamentals or are ready to learn to write your own Neural Networks, Brilliant has a course for you: