In this digital age, where technology is advancing rapidly, there’s a sneaky player in town – deepfake videos. A deepfake is a fake video or audio created by powerful computer algorithms that can make someone appear to say or do things they never actually do. Deepfake tech is now near to perfect which makes it hard to tell what’s real and what’s not. These videos, created using sophisticated algorithms, can make it tough to distinguish between real and fake. I have already talked about Deepfake videos in one of my previous articles. Deepfakes are again in the news after a Deepfake video of Rashmika Bandanna is trending on social media. But this is not the only case. A Deepfake video or Katrina Kaif is also trending. We have already seen fabricated videos of political figures making controversial statements
Although deepfake of celebrities seems only for memes and entertainment purposes, Scammers are now using Deepfake to scam people. In a piece of recent news, an Ex-Coal India executive lost Rs 40,000 in Kerala. He got a video call from an old friend asking for money, while the video on the call for actually a deepfake video of his video.
Now that Deepfake is the center of several controversies, the Indian government has instructed social media platforms to remove deep fake videos within 24 hours of receiving a complaint. The government has also advised affected people to file FIR at the nearest police station and inform the social media platforms to take them down.
How to identify DeepFake videos
As a normal social media and internet user, you also need to know how to figure out if the video is real or deepfake to avoid the spread of misinformation. Deepfake technology has become good, but it is still not perfect. So, you can easily figure out if the video is Deepfake. You only need to have a closer look at a few things. In this detailed article, I will tell you how to spot Deepfake videos.
Watch those Faces: Keep a close eye on facial expressions. Deepfakes often give themselves away with awkward or unnatural movements. If the lighting on the face seems off or if things look blurry, that might be a red flag.
Eyes Don’t Lie: Look into the eyes. Deepfake eyes may not blink naturally, and reflections might not match the surroundings. It’s like a game of hide-and-seek but with reflections.
Fine Details Matter: Real faces have pores, wrinkles, and imperfections. Deepfakes might miss these fine details. If the skin looks too smooth and perfect, be skeptical. If you suspect there’s a mole, check if it looks genuine. Typically, DeepFakes struggles to make changes to facial hair look completely natural.
Lighting Woes: Check for consistent lighting and shadows. Deepfakes struggle to maintain harmony between the manipulated face and the original background. Shadows acting funny? That’s a clue.
Head Scratcher: Watch out for head movements that don’t match the body or the background. If the head is doing a dance of its own, it might be a deepfake party.
Listen Up: Pay attention to the audio. Sometimes, the voice doesn’t sync with the lip movements. If it sounds like a bad dub, it might just be one!
File Sherlock: Dig into the file’s metadata. Any unusual info lurking there? But remember, sneaky creators can tamper with metadata, so don’t rely on it alone.
Deepfake Busters: Lucky for us, there are tools out there designed to sniff out deepfakes. Microsoft Video Authenticator and Deepware Scanner are like digital detectives, always on the lookout.
Backtrack with Search: Use reverse image and video searches. Has the content been seen before? Deepfakes often recycle existing footage, so a quick search might unveil the truth.
Common Sense Check: Lastly, trust your instincts. Does the video make sense in the given situation? Context matters, and if something feels off, it probably is.
Wrap Up
It is possible to detect deepfakes, but is now easy. I have already explained different things you need to consider to spot Deepfake. It’s essential to stay updated on new developments in deepfake technology and detection methods, as the landscape is constantly evolving. Keep in mind that no method is foolproof, and a combination of these approaches may enhance the chances of identifying a deepfake video.
