I bumped into this youtube channel where I'd found a plethora of cool topics, deep learning, 3d graphics/simulations... the owner love for it also helped of course! Congratulations Károly Zsolnai-Fehér, I hope I've spelled your name correctly!
Just a few 2 minute talks amongst the many:
Adaptive Fluid Simulations
Adaptive algorithms are a class of techniques that try to adapt at the problem that we have at hand. This adaptive method focuses computational resources to regions which are visible and have many fine details, and coarsens the simulation quality in regions that are not visible (or interesting).
Modeling Colliding and Merging Fluids
Surface tracking is required to account for topological changes when different fluid interfaces collide, taking into consideration the possibility of colliding fluids that are made of different materials.
Narrow Band Liquid Simulations
We continue our journey in the land of fluid simulations and discuss a really cool FLIP-based technique that uses both particles and grids to create very high quality footage at a much more reasonable cost than previous works.
Fluid Simulations with Blender and Wavelet Turbulence
Wavelet Turbulence is a technique that helps achieving fluid and smoke simulations, but orders of magnitude faster. It is also much lighter on memory and is now widely used in the industry, and had given Theodore Kim an Academy Award for this SIGGRAPH publication.
Simulating Breaking Glass
This new technique enables the simulation of tearing for a variety of materials like cork, foils, metals, vinyl and it also yields physically correct results for glass. The algorithm also lets artists influence the outcome to be in line with their artistic visions.
Painting with Fluid Simulations
As there is a lot of progress in simulating the motion of fluids, and paint is a fluid, then why not simulate the process of painting on a canvas? The simulations with this technique are so detailed that even the bristle interactions are taken into consideration, therefore one can capture artistic brush stroke effects like stabbing. Traditional techniques cannot even come close to simulating such sophisticated effects.
Artistic Manipulation of Caustics
A caustic is a beautiful phenomenon in nature where curved surfaces reflect or refract light, thereby concentrating it to a relatively small area. Since we, humans are pretty bad at estimating how exactly caustics should look like, one can manipulate them to be more in line with their artistic vision.
3D Printing Objects With Caustics
What are caustics? A caustic is a beautiful phenomenon in nature where curved surfaces reflect or refract light, thereby concentrating it to a relatively small area. This technique makes it possible to essentially imagine any kind of caustic pattern, and it will create the model that will cast caustics that look exactly like that.
Interactive Editing of Subsurface Scattering
Subsurface scattering is a technique to model light transport not between surfaces, but volumes - it therefore enables rendering digital images of human skin, marble, milk, and many other translucent materials. This piece of work takes this a step beyond, and offers a compelling solution to editing subsurface scattering.
Separable Subsurface Scattering
Separable Subsurface Scattering is a novel technique to add real-time subsurface light transport calculations for computer games and other real-time applications.
From Doodles To Paintings With Deep Learning
This technique uses deep learning to create beautiful paintings from terribly drawn sketches. The results look so great that many people called this work out to be an April Fools' day joke!
Deep Learning Program Learns to Paint
Artificial neural networks have been shown to be apt at reproducing the artistic style of many famous painters, such as Vincent Van Gogh and Pablo Picasso among many others. All the user needs to do is provide an input photograph and a target image from which the artistic style will be learned.
Deep Neural Network Learns Van Gogh's Art
This time they have been shown to be apt at reproducing the artistic style of many famous painters, such as Vincent Van Gogh and Pablo Picasso among many others. All the user needs to do is provide an input photograph and a target image from which the artistic style will be learned.
Artistic Style Transfer For Videos
This time they have been shown to be apt at reproducing the artistic style of many famous painters, such as Vincent Van Gogh and Pablo Picasso among many others. All the user needs to do is provide an input photograph and a target image from which the artistic style will be learned.
Randomness and Bell's Inequality
What makes an event random, and how incredible it is what Bell's theorem (or inequality) has to say about truly random events.
Capturing Waves of Light With Femto-photography
In their newest work researchers at MIT and the University of Zaragoza managed to to capture how waves of light propagate in space. We can observe how a mirror lights up with its image as light propagates from the light source to the camera.
Designing Cities and Furnitures With Machine Learning
Through Gaussian Process Regression, we get to populate a virtual city with buildings. All done in a very elegant and convenient way: it learns the preference of the user, then creates and recommends a set of solutions that are expected to be desirable.
Interactive Photo Recoloring
Image and color editing is an actively researched topic with really cool applications. Most of the existing solutions are either easy to use but lack in expressiveness, or too complex for novices to use. This algorithm try to preserve the full dynamic range of the input image image.
3D Depth From a Single Photograph
This piece of work tries to estimate depth information from an input photogaph. This means that it looks at the photo and tries to tell how far away parts of the image are from the camera, and the final goal is that we provide a photograph for which the depth information is completely unknown and we ask the algorithm to provide it for us.
Real-Time Shading With Area Light Sources
Shading with area lights adds a great deal of realism to CG renders. However, it requires solving spherical equations that make it challenging for real-time rendering. In this project, we develop a new spherical distribution that allows us to shade physically based materials with polygonal lights in real-time.
Face2Face: Real-Time Facial Reenactment
In computer animation, animating human faces is an art itself, but transferring expressions from one human to someone else is an even more complex task. This paper describes a real-time solution to this animation problem.
