Study Reveals Humans , May Not Be Able to Distinguish , Between Real and AI-Generated Faces.<br />PsyPost reports that new research was aimed<br />at determining how well the human brain can <br />detect AI-generated deepfake images. .<br />The teams' findings, published in 'Vision Research,' <br />suggest that the brain can distinguish subconsciously <br />distinguish between some deepfakes and real images.<br />The teams' findings, published in 'Vision Research,' <br />suggest that the brain can distinguish subconsciously <br />distinguish between some deepfakes and real images.<br />Throughout history, humans have <br />been regarded as the benchmark <br />for face detection. We have <br />consistently outperformed computers <br />in recognizing and classifying <br />faces (although this is changing), Mic Moshel, Study author & a PhD candidate in clinical <br />neuropsychology at Macquarie University, via PsyPost.<br />Throughout history, humans have <br />been regarded as the benchmark <br />for face detection. We have <br />consistently outperformed computers <br />in recognizing and classifying <br />faces (although this is changing), Mic Moshel, Study author & a PhD candidate in clinical <br />neuropsychology at Macquarie University, via PsyPost.<br />However, the emergence of AI has <br />presented a significant challenge <br />in reliably determining whether <br />a face is artificially generated. , Mic Moshel, Study author & a PhD candidate in clinical <br />neuropsychology at Macquarie University, via PsyPost.<br />Intrigued by this development, we sought <br />to investigate how humans respond to <br />hyper-realistic AI-generated faces, <br />specifically exploring the ability to <br />differentiate between real and fake, Mic Moshel, Study author & a PhD candidate in clinical <br />neuropsychology at Macquarie University, via PsyPost.<br />PsyPost reports that researchers used neural networks <br />called Generative Adversarial Networks (GANS) to <br />generate images of faces and everyday items.<br />EEG data was reportedly used <br />to analyze the brain's response <br />to real and fake images.<br />Our findings revealed that <br />individuals can potentially <br />recognize AI-generated faces <br />given only a brief glance. , Mic Moshel, Study author & a PhD candidate in clinical <br />neuropsychology at Macquarie University, via PsyPost.<br />Nevertheless, distinguishing genuine faces <br />from AI-generated ones proves to be more <br />challenging. Surprisingly, people frequently <br />exhibit the tendency to mistakenly <br />perceive AI-generated faces as <br />more authentic than real faces, Mic Moshel, Study author & a PhD candidate in clinical <br />neuropsychology at Macquarie University, via PsyPost.<br />PsyPost reports that the team's findings could <br />have implications for cybersecurity, counterfeiting, <br />border security and spreading misinformation.
