Periodic Reporting for period 2 - Talkitt (Speech recognition technology to enable people with speech disabilities to communicate freely)
Reporting period: 2018-07-01 to 2019-12-31
Speech is a basic need. There is an incredible, and simultaneously unaddressed, need for 21st century solutions that restore communication channels for all individuals. In the US and Europe, a combined 10.4 million people suffer from speech disabilities, and 8% of all children suffer from a consistent or temporary communications disorder.
The most common type of atypical acoustic control is dysarthria, which is caused by either congenital or acquired conditions that affect the physical production of speech. Congenital causes are usually the result of complications during pregnancy (like asphyxiation to the brain) — with cerebral palsy (CP) among the most common conditions. Around the world, CP is the most common cause of disability among children, and nearly 90% of all children with CP have a temporary or permanent diagnosis of dysarthria. For older individuals, late-onset dysarthria is most commonly caused by cerebrovascular stroke, which affects 1% of all individuals and 5% of people over the age of 65, globally. Other sources include amyotrophic lateral sclerosis (ALS), Parkinson’s Disease (PD), brain cancer, and traumatic brain injury (TBI).
Voiceitt is building speech recognition technology designed to understand non-standard and dysarthric speech. Our core mobile application – Talkitt - is a hands-free voice recognition app that enables face-to-face, real-time communication with friends, family, and strangers. In addition, our technology is designed for integration into smart homes, Assistive and Augmentative Communications devices, and smart speakers.
Voiceitt’s core mission is to make voice recognition technology truly accessible to everyone. Our principal aims are inclusion and independence, and we are committed to helping children and adults around the world communicate freely, spontaneously, and naturally by voice. Through a hybrid of unique statistical modelling and machine learning, Voiceitt will enable tens of millions of people to overcome communication barriers and help them connect with the world.
The most common type of atypical acoustic control is dysarthria, which is caused by either congenital or acquired conditions that affect the physical production of speech. Congenital causes are usually the result of complications during pregnancy (like asphyxiation to the brain) — with cerebral palsy (CP) among the most common conditions. Around the world, CP is the most common cause of disability among children, and nearly 90% of all children with CP have a temporary or permanent diagnosis of dysarthria. For older individuals, late-onset dysarthria is most commonly caused by cerebrovascular stroke, which affects 1% of all individuals and 5% of people over the age of 65, globally. Other sources include amyotrophic lateral sclerosis (ALS), Parkinson’s Disease (PD), brain cancer, and traumatic brain injury (TBI).
Voiceitt is building speech recognition technology designed to understand non-standard and dysarthric speech. Our core mobile application – Talkitt - is a hands-free voice recognition app that enables face-to-face, real-time communication with friends, family, and strangers. In addition, our technology is designed for integration into smart homes, Assistive and Augmentative Communications devices, and smart speakers.
Voiceitt’s core mission is to make voice recognition technology truly accessible to everyone. Our principal aims are inclusion and independence, and we are committed to helping children and adults around the world communicate freely, spontaneously, and naturally by voice. Through a hybrid of unique statistical modelling and machine learning, Voiceitt will enable tens of millions of people to overcome communication barriers and help them connect with the world.
Within the project, we have been actively working on the optimisation of the technology itself and have successfully launched the first app, available in four languages, for large scale beta testing. Beta testing was performed with leading institutions in Israel and across Europe, Ezer Mizion, Bambino Gesú, The Karten Network and Universitat Autonoma de Barcelona, who acted as both research and testing partners. The trials demonstrated that Talkitt represents an incredible breakthrough for the people involved and provides highly accurate and reliable performance.
We will continue to optimise our speech recognition technology following the successful results from the trials. The trials demonstrated that the social impact on the quality of life of the end users is highly positive and that almost all of the participants found new independence. As a continuation of the project, we will explore other ways of improving the technology and alternative ways in which it can be adopted e.g. integrating with a wearable device or smart speaker. The SME Instrument Phase II project has enabled the launch of the Voiceitt technology into the world's marketplace and enable individuals to communicate with their own voice. This project marked a major stepping stone in the development of the company, and for the development of the Assistive and Augmentative Communication (AAC) market as a whole. Not only are we enabling people to communicate with their own voices (a first for AAC technology), we are evaluating quality of life improvements in a clinical setting with UAB and Catalonia's leading disabilities organization, CEE Post del Drago.