Eyes tell all: How to tell that an AI generated a face?

Are you afraid you cannot tell Deepfakes apart from real images? What if we told you that there is a way to know whether a face was generated by an AI (at least for now, lol 🤖) ? Today we comment on the paper “Eyes Tell All: Irregular Pupil Shapes Reveal GAN-generated Faces.” We discuss the pupil quality of GAN generations. And more generally about cues on how to tell GAN-generated images apart from real ones. ➡️ AI Coffee Break Merch! 🛍️ https://aicoffeebreak.creator-spring.com/ 📺 GAN explained: https://youtu.be/_qB4B6ttXk8 📺 Visual Chirality (mirrored or not): https://youtu.be/rbg1Mdo2LZM Paper: 📜 Guo, Hui, Shu Hu, Xin Wang, Ming-Ching Chang, and Siwei Lyu. "Eyes Tell All: Irregular Pupil Shapes Reveal GAN-generated Faces." arXiv preprint arXiv:2109.00162 (2021). https://arxiv.org/abs/2109.00162 📎 Flickr-Faces-HQ Dataset (FFHQ) https://github.com/NVlabs/ffhq-dataset 📎 Browse faces generated by StyleGAN2: https://thispersondoesnotexist.com/ Outline: 00:00 Eyes tell all 01:57 Are all GANs affected? 03:47 Where to look for errors? 06:03 Bias Thanks to our Patrons who support us in Tier 2, 3, 4: 🙏 donor, Dres. Trost GbR, Yannik Schneider ▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀ 🔥 Optionally, pay us a coffee to help with our Coffee Bean production! ☕ Patreon: https://www.patreon.com/AICoffeeBreak Ko-fi: https://ko-fi.com/aicoffeebreak ▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀ 🔗 Links: AICoffeeBreakQuiz: https://www.youtube.com/c/AICoffeeBreak/community Twitter: https://twitter.com/AICoffeeBreak Reddit: https://www.reddit.com/r/AICoffeeBreak/ YouTube: https://www.youtube.com/AICoffeeBreak #AICoffeeBreak #MsCoffeeBean #MachineLearning #AI #research​