ML FOR INNOVATIVE ASSISTIVE TECHNOLOGY
Competentum has made significant progress in the use of machine learning (ML) and artificial neural networks (ANN) for the most accurate speech and image recognition. This expertise enabled us to create two innovative assistive technologies tailored to inclusive online education – Voice ART and Hand ART.
Assistive Technology for Accessibility & Assessment
Speech and image recognition technology removes barriers in education for people with different disabilities and health restrictions (visual impairments, motor skills etc.) as well as provides new learning opportunities for people without reading and writing skills (preschoolers, immigrants), when using AR/VR devices, or working on the complex STEM tasks with formulas and graphs.
Voice ART for Text-To-Speech & Voice Recognition
Voice ART is a voice controller that works as a standalone tool or as a SaaS solution integrated into various LMS. The main benefits of using Voice ART:
- Language models for Voice ART are built based on the linguistic corpora for eLearning. This architecture allows the tool to recognize speech in the context of eLearning more accurately compared to the voice assistants for general purposes.
- Voice ART has the ability to manage navigation both inside a page and between pages, as opposed to the existing speech recognizers, which can only switch between browser pages and cannot perform pedagogically meaningful actions. Considering the variety of ways to enter response variants in online tests (drag-and-drop, formulas and graphics, etc.), this functionality is crucial.
- Voice ART makes it possible for students to enter mathematical and chemical formulas using their voice, converting the input into Latex and MathML format.
Hand ART for Handwriting Recognition
Hand ART is an innovative handwriting recognition tool that converts the conventional form of writing into a digital testing environment. A student can use it with a finger or stylus on any device, system, or LMS. Hand ART recognizes the handwriting on a photo or by using electronic ink and adapts the written expression to the problem type. After submitting the answer, instant feedback is provided. Hand ART shows the full and final picture to sum up and evaluate the work. Hand ART can be easily integrated into any LMS or other digital assessment software.