Category Archives: Uncategorized

Mathematics – The magic behind animated movies

Nowadays, most animated movies are made by using computer animation techniques instead of hand-drawn animation. When we watch these computer animation movies, we usually think the producers are artists with a creative mind and great drawing skills. However, not many of us have noticed that some of them are also great mathematicians. So the question is how maths is used in computer animation?

Creating an animated movie is like creating a new virtual world, where new planes need to be defined in a coordinate system. A plane is described by a set of three points or a point and 2 vectors. But imagine a world where everything is on a flat infinite plane, well, this must be a boring world. Therefore, hills, mountains, rivers, valleys are added. It sounds like an easy task, but in fact it isn’t easy at all. First, the objects are sketched by using wire skeletons which are composed by simple polygons. A mathematical methods called Non-uniform rational basis spline (NURBS) is then used to smoothen surfaces of the objects. In a TED-ed talk, Tony DeRose, a Senior Scientist and Research Group Lead at Pixar, gave a presentation about how Non-uniform rational basis spline (NURBS) was used in Pixar’s animated movies. In a paper called Subdivision Surfaces in Character Animation, which was published in 1998 by Tony DeRose, Michael Kass and Tien Truong, the application of NURBS in creating subdivision surfaces was described in detail.

YouTube Preview Image

Now it’s when coordinate systems get involved. As the objects are put together, coordinates are required to make sure the objects locate at where they’re supposed to be. Vectors are used to define the location of every point with respect to the origin of the chosen coordinate system. To describe the motion of an object in space, transformation matrices are used. Depending on the elements of the matrices, the transformation can be rotation, translation, enlargement, reflection or a combination of these.

 

Some examples of simple transformation matrices

The following video shows how maths is use to create more realistic animations

YouTube Preview Image

The above mathematical methods are only a few main methods that are used in computer animation. There are many other mathematical methods that are useful for computer animation, such as Fourier Transformation, Complex Analysis, and so on. Mathematics plays an important role in computer animation. It makes a significant contribution to the success of computer animated movies.

Ryan Tran.
Continue reading

Self-Driving Cars: The future of transportation?

Vehicles, trains, ferries, planes – just examples of the wide range of transportation methods available in the modern day.  All of these methods require a human operator, however, we are soon to be introduced to a method seldom heard of: self-driving cars.  In 2009, Google launched the Self-Driving Car Project and began development of a vehicle that drives itself from point A to point B without the need of a driver.

Google_self_driving_car

Image: Wikimedia (by Michael Shick);  Google Self-Driving Car in Mountain View, CA, USA.

An operating software works alongside the sensor seen visibly on top of the vehicle. The sensor consists of lasers and cameras, and it is able to differentiate surrounding objects by a combination of size, speed and shape. With this information, the software is able to anticipate what happens next (for example, if a cyclist passes by).  In the current prototype, the car has no pedals nor steering wheel – the only way humans can control the driving is a red button for emergencies. Google’s prototypes have been released in California and are expected to be released to the public by 2020.

YouTube Preview Image

Other examples of  self-driving vehicles include Tesla models, which have been on the roads of Canada.  Compared to Google’s project, Tesla’s models are not entirely automatic, as they still require a driver to command lane changes and they require a driver to be present and touching the steering wheel periodically.  Nonetheless,  it still contains functions such as cruise control, automatic parking,

2013_Tesla_Model_S_(11322176214)_cropped

Image: Wikimedia (by Niels de Wit);  Tesla Model S, first Tesla to have autopilot software.

hand-free lane changes, and top speed limit.

Apart from not needing a driver, what else does a self-driving car offer?  Google’s prototype is environmentally friendly because it is completely battery powered. The computer of the vehicle is also programmed to drive safely, although Google has the ability to set the aggressiveness of the drive.  Consultancy firm KPMC predicts self-driving cars can reduce accidents by 80% and an estimated 2500 fewer deaths due to car accidents between 2014 to 2040.

The self-driving car still has it’s limitations though.   The software in all models are still to be perfected and we are still seeing accidents in prototypes. Questions have also been raised about common driving circumstances,  such us backing up a few feet for convenience, or deciding if someone should be let into your lane. If every car on the road was programmed, perhaps we could little to no car accidents.  However both the hardware and software is fairly expensive, such as Tesla’s software requiring a $3000 add on.  In addition there are various legal issues. Ontario still requires a driver  to be in the driver’s seat at all times. Responsibility of accidents are still to be decided between the user and manufacture.

With many factors still unknown, the popularity of the self-driving car in the future is still debatable. However, we can continue to expect steps in development and new self-driving models to appear.

Daryl

 

Howdy, Dr. Robot?!

Human health…. People spend tons of money to maintain

image from pixabay.com

Image from pixabay.com

their health in a good condition and even more when something goes wrong. Therefore, one would expect the accurate diagnosis and adequate treatment from the cohort of specialists available on the market nowadays, especially when it comes to the early stages of the most hardly detectable, but lethal diseases, like cancer. Till the latest time, there was no other alternative for the patients than to throw themselves into the hands of the most skillful, knowledgable, and experienced medical specialists in the field. But is it all shine and no rain?

According to a recent research, medical errors are on the third place among the all fatality causes in medical community. Among the main reasons are miscommunication, inability to keep track on the patients’s extensive medical history and history of drug consumption, and misreading the data patterns received from the tests. The latest peak of notorious cases occurred in Newfoundland and Labrador, Canada, and related to the doctors who continuously misdiagnosed their patients with breast cancer and prescribed the unnecessary treatment or no treatment at all, when it was needed. The biggest issue there was lack of “quality control”, so, no one questioned or reviewed the diagnostic processes!

At the same time, the progress in Artificial Intelligence and Machine Learning brought to life several brilliant projects, which have a better accuracy in diagnosing patients than the doctors have. For instance, the IBM Watson Health system gathers different types of information, including the recognition of the medical images, into one immense database. From that, the “self-learning” analytical technology recognizes the patterns and makes more accurate predictions. In general, the machines are better in interpreting the patients’ symptoms and medical history. And it costs less money!

More on how the machines may “outsmart” people in the video below (Herbert Chase, MD, MA on TEDMED Day CUMC, Youtube).

Among the benefits of having machines, if not as “doctors”, but at least as the doctor’s “assistants”, are the better interpretation of the clinical tests, control over the prescribed treatment, ability to fast-track the possible negative reactions or, so-called, adverse effects, and creating a continuous medical history, which is easy to review. And that’s what the scientists at MIT Artificial Intelligence lab work at.

The only possible draw back is creating a sufficient “learning pool” for the

Bodymedia device collects data about the person’s activity and burnt calories. Image from flickr.com.

Artificial Intellect, as an enormous amount of data is required to predict the most possible outcomes. However, as the technological progress moves on, the problem of data collecting becomes less and less significant, as the personal wearable devices  open the new opportunities in this case.

As seen from above, new technologies are rapidly moving into the health industry. I hope, soon we all can step into the era where doctors and machines will work together for the benefit of society and where it will be no place for the medical errors.

~ Alex Budkina