Citlalli Selene Avalos Montiel and Flor Yelitze Pérez García of the IPN Engineering and Advanced Technology Interdisciplinary Professional Group have developed a prototype that can help people with language barriers integrate.
Mexico City (apro). -Graduates of the National Institute of Technology (IPN) created a prototype glove with a sensor that translates speech and text in Mexican Sign Language on a mobile device.
??Speaking hand?? is a prototype of Citlalli Selene Avalos Montiel and Flor Yelitze Pérez García. The prototype comes from the Interdisciplinary Professional Department of Engineering and Advanced Technology (UPIITA), which uses two pairs of gloves to detect Mexican sign language and send it to mobile devices. It also comes with the Android operating system, and the Translator application has been pre-installed before to display letters one by one until a word is formed, and these words can also be heard.
?? It is easy to register a glove device on your mobile phone. First you must find and select “GloveDer” in the “Bluetooth” section of “Settings”, and then in the application menu, select the translator with two white hands Icon background and pink outline??, Telematics engineer explained.
The prototype is composed of a pair of gloves. Each glove has a flexible sensor and accelerometer that can stretch the curvature of the finger. It also has a switch designed by the Polytechnic Institute, which can be adjusted according to the joints of the hand, and a data is also elaborately produced. The acquisition board communicates via Bluetooth, and they also use wires and flexible cables.
Teachers Miguel Félix Mata Rivera and Susana Araceli Sánchez Nájera from UPIITA made suggestions. The young polytechnic said that they had adopted another prototype developed in 2015 as the basis for their project, but this time, due to advances in artificial intelligence, they used artificial nerves. The network can improve the recognition and be trained to detect the symbol of the letter ?? A ??. To ?? O ??.
To achieve a neural network that accurately detects each executed gesture, extensive training and up to 15 or more repetitive positive reinforcements are required. This is like when we teach a child the name of an object, we will Repeat it over and over again, but the engineer commented, is it more effective when it is accompanied by feedback?
The polytechnic recognized that due to the wear and tear of glove materials and data acquisition cards, apart from the COVID-19 pandemic, the remaining letters have yet to be trained. However, improvements to the prototype still exist as a basis for continuing to explore the various possibilities of artificial intelligence .
For telematics engineers, sign language is similar to any other foreign language. This is a must to know so that people with hearing and speech impairments can feel integrated into society. Therefore, the idea of developing this device is to use the development of technology so that anyone has the possibility of communicating and facilitating their daily lives.