Charmin creates AI digital doppelganger for video calls
Charmin’s latest prototype is a video chat bot called BRB Bot. It is the ‘first-ever bot’ that will keep you logged on to a video call by creating a digital twin that swaps your live video feed with an AI-powered version of yourself.
BRB Bot, which is currently beta-only, is a desktop app that accesses the camera feed and uses machine learning, natural language processing, and tone analysis to listen to video calls and serve up the appropriate reactions in real-time so no one notices you’ve left to take a break.
According to the BRB Bot listens and responds as if you are still on the video call and features a range of realistic human reactions, such as laughing at your boss’ terrible jokes, looking thoughtfully into the distance, pretending to search for the unmute button and much more. Before using it, you have to record those clips through the software, responding to a variety of prompts with the appropriate action and expression so the bot can cut those clips in later. It remixes clips of your face to give the impression that you're still in front of the camera even when you step away.
“From video conference calls, to gaming, to sports, we’ve been forced to adapt to living and working virtually; however, one thing that hasn’t changed is when nature calls, you have to answer… no ifs, ands or butts about it,” said Rob Reinerman, Charmin Vice President, Procter & Gamble. “At Charmin, we’re obsessed with delivering a better bathroom experience whether it be providing the best toilet paper or inventing other novel ways to help people Enjoy the Go. BRB Bot is yet another way Charmin is exploring better bathroom technology that could one day become reality.”
Risks of AI bots
Although there can be benefits of AI bots that enable you to be gone for a minute or two on a video call without being noticed, it begs the question of how far people could take it.
The COVID-19 pandemic has meant a lot more video meetings than usual. Bored with having to be on so many Zoom calls, creative technologist decided to build an AI-powered clone named ‘Zoombot’ to handle the meetings for him. Reed designed Zoombot to use speech recognition and text-to-speech to respond to his colleagues and the video rotates through several images of Reed in front of his computer, making it seem like he has a bad connection.
With advancements in technology, these AI bots could enable people to skip meetings, miss important information, and not be present at all, as long as the bot is realistic enough. The technology needs to be used in the right way and not abused by people who wish not to attend meetings all the time.
Facebook Develops AI to Crackdown on Deepfakes
In light of the large tidal wave of increasingly believable deepfake images and videos that have been hitting the feeds of every major social media and news outlet in recent years, global organisations have started to consider the risk factor behind them. While the majority of deepfakes are created purely for amusement, their increasing sophistication is leading to a very simple question: What happens when a deepfake is produced not for amusement, but for malicious intent on a grander scale?
Yesterday, Facebook revealed that it was also concerned by that very question and that it had decided to take a stand against deepfakes. In partnership with Michigan State University, the social media giant presented “a research method of detecting and attributing deepfakes that relies on reverse engineering from a single AI-generated image to the generative model used to produce it.”
The promise is that Facebook’s method will facilitate deepfake detection and tracing in real-world settings, where the deepfake image itself is often the only information detectors have to work with.
Why Reverse Engineering?
Right now, researchers identify deepfakes through two primary methods: detection, which distinguishes between real and deepfake images, and image attribution, which identifies whether the image was generated using one of the AI’s training models. But generative photo techniques have advanced in scale and sophistication over the past few years, and the old strategies are no longer sufficient.
First, there are only so many images presented in AI training. If the deepfake was generated by an unknown, alternative model, even artificial intelligence won’t be able to spot it—at least, until now. Reverse engineering, common practice in machine learning (ML), can uncover unique patterns left by the generating model, regardless of whether it was included in the AI’s training set. This helps discover coordinated deepfake attacks or other instances in which multiple deepfakes come from the same source.
How It Works
Before we could use deep learning to generate images, criminals and other ill-intentioned actors had a limited amount of options. Cameras only had so many tools at their disposal, and most researchers could easily identify certain makes and models. But deep learning has ushered in an age of endless options, and as a result, it’s grown increasingly difficult to identify deepfakes.
To counteract this, Facebook ran deepfakes through a fingerprint estimation network (FEN) to estimate some of their details. Fingerprints are essentially patterns left on an image due to manufacturing imperfections, and they help identify where the image came from. By evaluating the fingerprint magnitude, repetition frequency, and symmetrical frequency, Facebook then applied those constraints to predict the model’s hyperparameters.
What are hyperparameters? If you imagine a generative model as a car, hyperparameters are similar to the engine components: certain properties that distinguish your fancy automobile from others on the market. ‘Our reverse engineering technique is somewhat like recognising [the engine] components of a car based on how it sounds’, Facebook explained, ‘even if this is a new car we’ve never heard of before’.
What Did They Find?
‘On standard benchmarks, we get state-of-the-art results’, said Facebook research lead Tal Hassner. Facebook added that the fingerprint estimation network (FEN) method can be used for not only model parsing, but detection and image attribution. While this research is the first of its kind, making it difficult to assess the results, the future looks promising.
Facebook’s AI will introduce model parsing for real-world applications, increasing our understanding of deepfake detection. As cybersecurity attacks proliferate, and generative AI falls into the hands of those who would do us harm, this method could help the ‘good guys’ stay one step ahead. As Hassner explained: ‘This is a cat-and-mouse game, and it continues to be a cat-and-mouse game’.