Experiment

 

AR Jutsu brings you to the Shinobi World and allows you to play with different jutsus in AR. Choose your favourite jutsu and bring it to live!

Rasengan

Rasengan is a spinning ball of chakra that can be held by the user. With its power, it can burst away a solid wall or an opponent.

This jutsu is achieved by detecting the hand region and placing the spinning ball in the center of the palm. The ball always remains at the center of the hand as it is tracked by the real-time hand tracking algorithm.

Swirls are triggered when the hand collides with a wall. This is done by using plane detection to determine all the available planes in the scene. Once the hand is close to any vertical plane enough, a custom swirl shader is applied on that particular mesh.

Summoning Jutsu

Summoning Jutsu allows the user to summon an animal to their location. To summon, the user has to place a hand on the ground and channel chakra to that hand.

Here too, the hand and plane detections are used. A summoning seal appears in the scene once the user’s hand collides with a plane, followed by a smoke effect and finally the 3D model of the summoned animal, in this case, a nine-tailed fox.

Before the AR Jutsu, people had to rely on computer graphics or video editing to achieve a real-life jutsu. Creating a context-aware AR app by combining AR with Machine Learning opens up new aspects to blend reality with the fictional world.

Now, since we’ve already handled the complicated parts, the fans just need to add their creativity to bring their beloved Naruto’s universe to life, in real-time. No post-production required.

If you want us to create any jutsu, let us know!

*Disclaimer* – This is a Naruto fan made app, for entertainment purposes only. No copyright infringement intended. All rights credited to the owner.

Geometric Clock
Geometric Clock is our experiment on a digital clock that tell times using simple 3D shapes. Number of shapes on the screen tell how much time has passed. Three columns are used for hours, minutes and seconds.

Geometric Weather Report
Geometric Weather Report is an installation that uses 3D cylinders to report wind direction in a stylish way. Cylinders are formed as a flow field that bends with the flow of air. The direction of each cylinder tells the direction the wind is blowing. Each column of the cylinders represent one timestep of 15 minutes. We can report both wind direction in the past and the predicted future direction with this installation.

Circular Wave is a computational art algorithm we designed inspired by the form of waves. To represent wave’s boundless property, we enclosed them in a circle form. With various adjustable parameters, countless beautiful art pieces can be obtained.

Dynamic Alphabet is our experiment on geometric objects inserted into the space within hollow alphabet. Squares and circles in this dynamic computational art randomly change their sizes and we let physics rule adjust their positions, creating controlled yet unexpected effects on the letters.

Is it possible for machines to feel emotions? As we know for now, machines have no emotion. But if we humans teach them, will they be able to learn? (A)I Feel is a project dedicated to find answers to our questions by creating a teaching & learning process between humans and a machine. To teach the machine, it asks each user to draw a picture which represents a specific emotion. The picture will be recognized then memorized as one of representatives of the emotion. Then, to demonstrate what the machine has learned, the user can draw pictures to portray his/her emotion at the moment. The machine will interpret user’s emotion and colors the picture accordingly. The whole process will be visualized as an interactive installation which encourages people to participate in the project.

We present a shadow play system that displays in real-time the extrapolated motion sequences of the still shadow images casted by players. The crux of our system is a sequential version of the recently developed generative adversarial neural networks with a stack of LSTMs as the sequence controller. We train the model with preprocessed image sequences of diverse shadow figures’ motions, and use it in our enhanced shadow play system to automatically animate the shadows as if they were alive.

bit.bot is our latest experiment by our interns to create an adorable robot that can assist human withs various tasks such as drawing, cooking, or searching videos online. bit.bot is made with raspberry pi equipped with a tiny screen to let it exhibit various facial expressions, and a pico projector to display various videos to humans.
After saying the hotword “bitbot”, human can make queries which are then passed through the AI system. bit.bot then exhibits various facial expressions through the tiny screen and the result videos accordingly.