What you get:
10 ideas for machine learning projects.
Free datasets to use for your projects.
Examples of libraries and algorithms for each case.
Samsamia Technologies is a startup that applies robotics to the fashion market. The goal of robotics is to create autonomous robots that help us in tasks we do not want to perform, like walking the dog, ironing, wash the dishes, ironing, make the bed and some other dangerous tasks like ironing.
Robotics is a very young science, we could say that it begun in the 1970s, and was born as mix of another three sciences: mechanics, electronics and computer science. As a baby branch of science, it has many challenges to solve. Right now we are very far from what we see in movies and it will pass many years until we could have in our houses humanoid robots like HOAP-3 helping us.
Due to the difficulty of the problems robotic scientists address, it is usually divided in small parts. It is not easy to distinguish between these parts since many of them are very similar.
A very popular part is the Artificial Intelligence or AI. AI has to do with taking decisions.
Things like how can I go from the room I am right now to another room or which tool should I use if I need to clean a room.
All these actions require a high capacity of learning and a deep understanding of the environment. The ability to learn is studied in Machine Learning and the ability to understand the environment is studied in Pattern Recognition.
Machine Learning tries to imitate the amazing human capacity of learning. We are learning machines and our immense capacity to adapt to different situations has allowed us to evolve to what we are right now. The current capacity of robots to learn is very limited. However, some awesome advances are being made in the last years. A research performed by the CEO of Samsamia and scientists of Universidad Carlos III de Madrid and King’s College of London made a humanoid robot to imitate a behavior demonstrated by a human teacher and improve the learned behavior just like we human beings do.
Pattern recognition has to do with the ability of generalizing. For us it is easy to recognize a glass and that we usually use it to drink. If we don’t have a glass, we can even use a bottle to drink or any other recipient. But there are a million shapes of glasses or bottles and for a robot it is very difficult to recognize a glass, a cup or a bottle and that all of them are recipients and can be used to drink. There are many applications of pattern recognition, like fashion recognition, object and face detection and even emotion recognition. A cool development that our CEO, Miguel González-Fierro, did was a system that recognized not only the face of a person but also the emotion that this person is feeling, allowing a humanoid robot to reply in accordance.
Robotics is an amazing science and it has awesome applications. In Samsamia, we desire to show people how robotics, technology and science can be used in their life.