When released back in 2017 for iOS and Android devices, this AI-based face changing/photo editing app — FaceApp took the internet by storm. And with its mind-boggling results, it also created a major kerfuffle on social media platforms like Facebook, Twitter and Instagram, where everybody including celebrities was using the app to transform their selfies to see their “older version.” At the inception, the app was renowned for its ethnicity filters, but later it evolved with a lot of features.
However, the trend of transforming selfies into older self faded off as it started to raise some serious privacy and bias concerns, where the company had access to those millions of photos and shared the data to their servers. In fact, last year, the FBI issued a letter stating how these Russian apps can be a potential threat to the country.
However, built by a Russian startup — Wireless Lab, after almost three years again exploded on social media platforms with its new gender swap feature that allows users to transform their selfies and see how they would look with a different gender. Even Snapchat last year released a similar gender swap feature, which exploded on the internet with its uncanniness.
In India, this trend of gender-swapping again started to catch up with cricketer Yuvraj Singh transforming the entire cricket team into women on his Instagram feed. The app claims to use AI to transform the selfies where the system analyses the photo and then transforms it according to its understanding. Although the company cleared many doubts about their privacy polls last year, however as the system uses artificial intelligence to analyse the faces of the users, it creates a lot of doubts.
In fact, Fabio Assolini, a senior security analyst at Kaspersky stated to the media warning users about the face app, as with recent developments, the facial recognition technology is mainly used for authentication password-protected access, and therefore users sharing data to third-party apps could make them vulnerable to hacker attacks. He also cautioned users about privacy concerns where these companies can share and sell these images for illegal purposes such as face modification, deep fake generation, etc. “… data is stored on third-party servers, and that it can also be stolen by cybercriminals and used to impersonate identities,” said Assolini.
Furthermore, in another news article, it has been argued that how these gender swap apps became a tool of joke on the internet, where people have gender-swapped Elon Musk’s image and shared it on Reddit. To which, Charlie Knight, a non-binary editor and activist, said to the media that, “Exploring gender is a good thing, I encourage it. You may learn new things about yourself that surprise you, and you may find yourself kinder to trans people. But in this context, it’s made into a joke, and that bleeds into the way most of society – most of the people using these filters – consider our transitions a joke.” Thus, it also becomes offensive to transgender people and triggering gender dysphoria among people.
Countering these allegations, the CEO and founder of the app, Yaroslav Goncharov told the media that the company deletes its users’ photos within 24-48 hours of their use. “We do not use the photos for any reason other than to provide the editing functionality,” said Goncharov.
However, such concerns always raise an interesting question of what is the tech behind this Russian app, and how artificial intelligence works to produce such realistic photos of the users.
How Does Artificial Intelligence Work In FaceApp?
The FaceApp is an image manipulation app that allows its users to alter and transform their images using filters. For this the app works on image recognition technology, that is key for facial recognition systems, and utilises deep learning for recognising the key features, like eyelids, cheekbones, jawline, nose bridge etc. of the human face to create those transformations.
Like any other machine learning model, this app also works on sample data that is usually gathered from the users’ mobile. Once the sample data, which includes pictures of the users, family members, friends and everything else, is collected, the system then provides the data to the deep neural networks of the app which helps the system to learn all the nitty-gritty of the human face. Considering deep learning models work on sample data, the more the users share the images as sample data, the more accurate the system is going to predict those transformed images. This predictive ability allows the app to simulate wrinkles, enhance receding hairline, and shrink the skin at a realistic level.
However, this app is entirely different from Instagram filters, as it uses AI technology to bring modification on the face. And to create these images, the app uses deep generative convolutional neural networks — a powerful group of networks — which is typically used for high-fidelity natural image synthesis, augmenting data, and enhancing image compression. These GANs are a structure of an algorithm that uses two conflicting neural networks contesting against each other to produce new information.
The GANs designed to recognise patterns, are usually unsupervised and learn on their own to imitate images according to the information based on the data fed into it. And therefore, the app takes facial information from one image and applies the same on the other image, and consequently, it works on the huge database it gathers from the millions of photos of the users.
Neural Network demonstration for Digit classifier // PC: blog.usejournal.com
However, by using GANs for facial feature alteration, it usually loses the real information and produces entirely new information for users. And to avoid this problem, the app uses conditional GAN with age parameter, which focuses on preserving the information they face, as the programmer can now control the output of the generator. Such conditions like age and gender can also be implied to both generator and discriminator networks while predicting outcome. Conditional GAN will also allow generating multi-modal models with varied conditions that can be applied like age, gender etc. So, for an age-specific filter, the app applies an added condition to the GAN, where it uses data that is only labelled by age. Similarly, for gender-specific filters, the app uses gender-labelled data.
The architecture of a conditional GAN
To make the prediction, the data goes through multiple players with neurons, and each neuron in the system have features, which are the functions of the sample data, and when the data is processed the output is sent to the next neuron with more complex functions, until the last neutron defines the output of the data to define the features of the face. Once the facial features have been identified, the system uses a generative adversarial network along with TensorFlow to apply the filters of age and gender.
To know more about GAN, click here.
The world has seen immense applications of GANs. However, this Russian app has revolutionised the technology and indeed has set a higher benchmark for photo manipulation apps in the market. Although the app comes with some privacy and security concerns, it indeed extended the use cases of GANs technology, which is now expected to be seen in more applications.