Machine learning, an interesting application of A.I.
What is machine learning? This is a very exciting application of artificial intelligence. In the traditional software development, experts tell the computer how to do a task. In machine learning, computers learn about the tasks, and when they have enough experience, they do it by themselves.
Machine learning vs. traditional programming
Machine learning is very important part of AI research and development. You can find a lot of solutions that are trained with the help of this technology. It can solve problems that traditional programming can’t.
In traditional programming, developers have to tell the computer how to do something, so you have to code the whole task. In machine learning, computers can learn automatically from datas. They need a lot of inputs, they looking for patterns in it, then they do the task automatically.
For example, if you want a song composed by an AI algorithm, you have to collect a lot of other songs in this style. You have to upload all of them. The computer learns like human do.
Learning computers need artificial neural network. Neural networks processing information. They are stimulating human brain, and consisting neurons. Neurons connecting to each other, like a web.
For example you want to compose a song with an AI algorithm. First, you have to train the system with the use of a lots of songs in similar style. That means a lot of data, you have to upload. The machine starts to find patterns in it. After it was trained, it starts to compose its own song.
Examples of using machine learning
As I mentioned before, with the help of machine learning we can make solutions that couldn’t be done with traditional programming. Speech recognition, self-driving cars, spam filters and movie subtitles are parts of our everyday life. These inventions are all developed with the help of M.L.
Besides this, I can also collect some surprising examples. Believe it or not, but an AI algorithm completed the unfinished symphony of Beethoven! Beethoven’s 250th birthday was in this April, and some musicians and programmers decided to finish the tenth symphony in this way. The team uploaded songs of Beethoven, and the computer started to learn.
If you love classical music, I have one further example for you. Chinese tech company Huawei’s AI smartphone finished the eighth symphony of Schubert. This solution was also trained with the help of M.L.
Because of M.L. computers become talented painters as well. I’m not talking about style transfer here. There is a new algorithm developed by researcher Mauro Martino. It goes one step further, because it needs only a picture of you. Then, the computer makes a perfect Renaissance portrait, just like it was painted by an old master.
If you are a big fan of history and technology, this one is for you. An AI algorithm colorized and revive an original vintage video filmed in 1911! It shows us the daily life of New York city.
The quality of this footage is excellent. It became slower, so it is like if we were just than and there. You can watch the full video if you click here. The original footage was shot by Svenska Biografteatern a Swedish newsreel company. The new, colorized version was made by Denis Shirayev, and the You Tube historian Guy Jones.
The Russian developer, Denis Malimonov created an AI that can take pixelated photo of a face and turn it to a lifelike one. This technology has some limits. Unfortunately, it can turn the photo in another face.
Our company, Netlife Robotics also use machine learning technologies. One of our product, Pepper robot (made by Softbank Robotics) can understand Hungarian language using our voicebot software.