Is it Possible to help deaf and dumb people to Express their thoughts, automatically translating , which they show in words? This question asked the Russian Junior Daniil Kazantsev from Yekaterinburg Lyceum. The solution to sort out this problem took Daniel to victory in one of the categories of the contest for young scientists Science Fair 2019. He developed a system that can interpret sign language and to replace a sign language interpreter.
As indicated , Kazantsev was selected among thousands of participants and became one of 24 finalists from 14 countries. In the end, he got the Lego Education Builder award for the development of his interpreter, sign language.
but All finalists. Daniel left
The basis for the development of the young Russian genius is the technology of electromyography, which determines the Electromechanical activity of the muscles of the changing position of the hands and fingers. In this case, each movement of the limb creates tiny electrical signals that can be read and then translate into understandable for the computer team. In particular, this technology is used to control artificial prostheses.
But Russian schoolboy has developed a technology based electromyography system which can translate the electrical signals of the muscles in speech.
a Blue and white crosses marked the location of the electrodes that read the electrical impulses of muscles
The Device consists of a cuff, which is worn on the forearm. Installed 24 on the cuff electrode, each of which reads the specific electrical signals of the muscles of the forearm. In addition, it uses a special sensor that determines the position of the forearm in space. On the basis of all these signals generate a unique "fingerprint" of the situation each individual part of the brush in space.
All collected signals are amplified and filtered from unnecessary electromagnetic noise, and then transmitted through the wires on the outside . The microcomputer digitizes the sensor signals and electrodes, and then transmits them into a powerful computing server with a specially trained neural network. With the help of the neural network is carried out interpretation (a translation word) data about the electrical signals that are responsible for each language gesture.
The system Check Daniel out on myself and two other volunteers. As the basis for the translation he used 5 gestures and popular sign language ASL (American Sign Language).
Hello (Hello); Yes (Yes); No (n) Please (please); I love you (I love you)
To verify the effectiveness of the system for each of the gesture recognition used 10 attempts. As the author points out the development, the system was able to show 92,6 percent accuracy in recognition.
but In the table indicates participants in the experiment, the names of the gestures that you want to recognize the system as well as the number of true positives
Live validation can be viewed below:
Daniel said, will continue work on Autonomous anoperation. There are a few things that do not yet allow to use it as a ready-made solution. For example, the system can translate only single gestures. But not yet able to form the received words of a sentence. So now the system can not be used as a complete replacement of the interpreter. The resolution of this issue will be one of the future tasks of the young scholar.
In addition, e-sign language interpreter requires regular access to a powerful computer server with a neural network that produces the signal recognition language gestures and translates them into words. However, the author notes that modern data transmission speed for mobile Internet allows permanent access to the server, thereby partially compensating for this deficiency. The author is also considering to make the device more compact, reducing the footprint of the electrodes to 7 centimeters.
In General, it is noted that in finalizing this technology, it can be used with any sign language dictionary. You do not need any special training to use this system.
More details about the development of Daniel Kazantsev see .
If you are interested in news of science and technologies, subscribe to our channel . There you will find exclusive content that was not published on the website!
Why is the new Oculus zuest 2 better than the old model? Let's work it out together. About a decade ago, major technology manufacturers introduced the first virtual reality helmets that were available to ordinary users. There were two ways to find yo...
DARPA has launched the development of a neural engineering system to research a technology that can turn soldiers into cyborgs Despite the fact that the first representatives of the species Homo Sapiens appeared on Earth about 300,000 - 200,000 years...
Marsha constructions n the surface of the Red Planet SpaceX CEO Elon Musk is hopeful that humans will go to Mars in the next ten years. Adapted for long flight ship Starship is already in development, but scientists have not yet decided where exactly...
At the moment the technology is already grown to the point when computer algorithms, along with special sensors that can read brain activity and use it as a control. Simply put, is the ability to create neural interfaces for . Sou...
the Sun is the source of existence of all life on our planet. Without the Sun, we would not have life. However, despite the fact that astrophysicists around the world have long been studying the structure of the main star in our p...
with the development of production of so-called "green energy" (solar and wind), the market is starting to need new solutions that can store this energy in large volumes. About two years ago, we wrote about how Tesla is actually o...