MIT is Training a New AI to Make Photographs Scary

MIT, Massachusetts Institute of Technology, is one of the leading universities that helps the world advance with its technology and now the scientists at MIT are training an AI to scare people!

This algorithm called “Nightmare Machine” is designed to turn a given image into a spooky one. It works like a scary filter and covers the image with frightening effects. The three people who wrote the software are sharing photos over the social media and asking people to choose which of them is the scariest one in order to train the AI and make it more terrifying. They are creating a learning mechanism by adding these data to the program and getting increasingly scarier photos.

This program which has been tested on celebrity faces and famous locations, achieves to give scary results.

Photo Application Plex Has Facial Recognition Through Artificial Intelligence

There have been recent important developments thanks to the AI applications built up by tech companies. Photograph and video editing software Plex uses a cloud based structure and combining it with AI, it can easily categorize your photos and videos by determining who they belong to.

Plex saves you from browsing through photos individually for hours by using its AI and indexing your thousand photo archives. Plex first analyzes your photos via AI technology and then identifies the objects in them. After identifying and naming the objects in your photos, Plex records these objects by using their names as tags. When you search for a photo by entering the name of an object in that photo, the application scans your archive to see which photos have that object and then show you the results.

A New AI Software Has Been Developed To Help Clear Low-Resolution Images

With a new software that has been developed by Google, low resolution images can be converted to more clear ones. Asides from increasing the resolution of the images, this software could also fill in the missing details. This way, by sharpening photographs with an 8×8 pixel quality, it will enable us to see unclear objects and faces.

This software can fill in the details by pairing a low-resolution image with less pixels and unclear details with a high-resolution image. This technology can also be used in photo editing and renewing old pictures.

Artificial Intelligence Can Detect Skin Cancer

Researchers from Stanford University in USA and Cancer Research UK in England, revealed that AI can detect skin cancer as well as a specialist via photographs. The AI that can detect skin cancer is developed based on the Google software which can recognize pictures of cats and dogs. After scanning 129.540 photographs with this software, it was observed that the type of skin cancer in the photos can de identified.

After that, the AI was taught to mark the most common skin cancer as “carcinoma” and the deadliest one as “melanoma.” Following that, the results were compared to the findings of 21 oncology experts.

As a result of this comparison, it was seen that the diagnosis accuracy of the AI was nearly equal to that of oncology experts’. The researchers of Stanford University stated that the outcome of this study is “exciting” and this AI will also be tested in clinical studies. Researchers of Cancer Research UK also stated that, this technology can be a quite useful tool for doctors. Experts say that the AI is a revolutionary invention for the health sector and everyone can turn their smartphones into a cancer detector.

The results of the study were published in “Nature” magazine.

Age Of Artificial Intelligence In Photo Editing

We are now in the time of AI in which the electronic devices have abilities like speech recognition, making a comment, bringing the closest results for your requests and analyzing them. Adobe is on a mission to make their products and software more useful with the help of AI. They are working on a kind of artificial intelligence that will turn Photoshop, which is already a successful photo editing software, into a program that can understand voice commands. Although this AI is currently in early stages of development, it is capable of executing standard functions on photographs. It can be quite useful when you want to edit a picture before sharing it on social media.

Adobe announced this project with a video. The program seen on the video has a microphone icon. This button activates the digital assistant and enables the user to cut, resize or share the project.

This video proves that something like this can be done. It is only a crazy concept project for now. Nevertheless, it is enough to get people excited.

8 Incredible Photos from Google’s Deep Dream

Deep Dream, Google’s photo-recognition tool, interprets and reproduces the images with its artificial intelligence algorithm. The millions of combinations within the program that engineers build with the so-called “artificial neural networks” offer unlimited imagination to the users.

Deep Dream

Deep Dream

Deep Dream

Deep Dream

Deep Dream

Deep Dream

Deep Dream

Photographs Will Take Up 75% Less Space Thanks To Artificial Intelligence

Google introduced a technology that decreases the bandwidth needed to show high quality photos and videos by 75%.

This new technology is called RAISR and it was first introduced in November. The name of this technique comes from Rapid and Accurate Image Super-Resolution and it uses machine learning to convert low quality images into high quality ones. It separates itself from the other super resolution methods on the market by being 10 to 100 times faster than them and thus being able to run in real time on mobile devices.

Google has been using this technology in various different services and now it is going to be used in the more common social service Google+. This technology enables 75% less bandwidth to be used in showing images on mobile and other types of devices. When Google+ users want to see an image, Google+ sends a request for one quarter of the pixels instead of the full version of the image. RAISR technology enables the same quality of an image to be produced by using less pixels.

This achievement of Google on images may be revolutionary for mobile bandwidth usage. Besides, the company partly gives an answer to Facebook that introduced its first mobile deep learning platform a while ago.

A little while before, Google announced that they succeeded in raising the energy use efficiency by 15% with deep learning company DeepMind. With savings like these, AI will help Google keep billions of dollars in the long run.

Artificial Intelligence Photography Studio

Intelligent robots may soon be among the world’s best fashion photographers. Dutch company StyleShoots introduced the smart studio system called StyleShoots Live. ‘Smart Studio’ can take photos and shoot videos at the same time while adjusting the technical aspects like lighting and background. The system is capable of editing videos in minutes with the help of an algorithm.

This product actually presents a detailed solution to a problem only a few of us are aware of. StyleShoots Product and Marketing Manager Anders Jorgensen says in his interview to Digital Trends: “We have identified some of the key problems of the brands wanting to do photo and video shoots of the models. Brands using different studios for stills and videos mean that twice the equipment, twice the space and twice the people will be needed. Besides, the models have a hard time changing studios between shoots. After the shoot, a group of people will have to go through and edit the images to produce high quality, clean clips.”

StyleShoots Live consists of a Canon DX Mk I camera and a camera head together with a computer engine that controls the lighting and all the other equipment. The stylist has the opportunity of using an interface to configure details like video styles and lighting settings. Jorgensen says: “All of these variables have infinite options and possibilities.”

StyleShoots Live is the latest example of AI’s capability of being applied to everything. However, the people working in creative professions do not need to worry yet. Jorgensen says: “In order to organize a fashion shoot, you need a creative eye to plan the shoot, pose the model and choose the clothing styles. A robot cannot do that – at least for now. So, we can say that our product does not replace the photographer. We are offering this product to people with a creative eye so that they can use their vision quickly and easily.

On the other hand, the set crew may lose their jobs as a result of this product. Jorgensen describes StyleShoots Live as “A cameraman, a crane operator and a video editor combined into one single machine.”

AI Can Find The People You Meet Outside, On Social Media

Shortly, we will be asking ourselves this: How can we prevent AI from learning about us? We upload so many information to the internet that the AI can pinpoint exactly who we are.

There is a huge photo sharing madness started with the ease of internet access and the widespread popularity of social networks. People either take selfies or group photos to upload them to social networks and share with their friends.

Naturally, these photographs do not vanish into the air or become forgotten. Asides from the conspiracy theories, the algorithms social networks and internet giants use, scan and analyze each photo shared. This way they can use these analyses for themselves in different services. For instance, a user can tag his friends in a picture automatically or apply the most suitable costume filters according to his features.

Russian photographer Egor Tsvetkov decided to take photos of the people he meets on the subway and try to find them on social networks. Egor carefully studied for 6 weeks and took photos of 100 people in the city of St. Petersburg. Then he analyzed them on the internet with a facial recognition software called FindFace.

The results were interesting. Egor chose VKontakte that was used in Russia as his preference of social network. He scanned this platform which has 55 million users and reached a success rate of 70 percent. That means he succeeded in finding 70 of the 100 people he took pictures of in the social networks. And this success rate is obtained only through VK.

Learning Abilities with Smart Cars and Artificial Intelligence

Sebastian Thrun, who is the leader of the team that won the smart car competition in 2005 DARPA Grand Challenge, talked in the documentary “Lo and Behold” and he wonderfully pointed out to an important learning ability difference between the AI and humans:

“When a smart car is involved in an accident and learns something from that, this learning experience is instantly shared with other smart cars. Thus, other smart cars learn from the same mistake too. This is something humans are not really good at. For instance, when I make a mistake – and I have made many driving mistakes in my life – I suffer from it and improve myself as a result. However, nobody else but me learns from that mistake. When a smart car makes a mistake, automatically all the other cars learn from that same mistake. Even the ones that are not manufactured yet! None of them will ever repeat the same mistake. This means that the cars’ ability to improve AI is far greater than humans’ capability to keep up with them.”

Probably in the future, cars driven by human guidance will be completely forbidden or unacceptable. The reason for this is not to restrict people; but the risk of having the control of a car being huge. Each of us will be able to do anything we wish to do while we are being transported to the place we need to go to.