VOICEITT BASIC

Speech-recognition and speech-generating solution.

Leading solution enabling people who have motor, speech, and language disorders to communicate using their own voice.

Release Date:
July 2017 (beta)

Voiceitt translates unintelligible sounds into clear speech in real time.


Voiceitt speech recognition technology translates non-standard speech patterns into clear speech in real time, enabling people with severe speech impairments to communicate spontaneously and naturally using their own voices.

Voiceitt, is easy-to-use and non-stigmatizing, and does not require any expensive device or technical expertise – disrupting the complex world of communication boards and tracking technologies to restore natural fluidity to millions of people with communication disorders.

FEATURES

The solution includes the powerful Motorola Moto G4 smartphone or LG Pad X tablet allowing the user to communicate freely with anyone, anywhere.

This innovative product is based on personalized speech recognition technology, which recognizes the user’s speech and translates it into understandable language that is displayed on and vocalized by the user’s device.

Voiceitt uses pattern classification technology, personalized for each individual speaker, is hands-free during use, and works in any language.

“In-time”: Voiceitt Basic allows for instant, real-time translation of incoherent speech patterns, allowing for more fluid, natural dialogue.

Completely hands-free / voice-activated – especially relevant, essential for users with motor control deficiencies.

Intuitive, simple interface: Voiceitt Basic does not require any technical expertise or programming; easy adoption by strangers or family members unfamiliar with AAC devices.

The on-device recognition, without reliance on continuous Internet access allows for nearly-natural speed by decreasing latency by at least 6X.

UNIQUE VALUE PROPOSITION
 
  • 1

    A voice-based intervention

    Voiceitt is the only Augmentative and Alternative Communication (AAC) solution that allows people with speech disabilities to communicate with their own voice as the input method.

  • 2

    Speech prediction

    By recognizing the unique phonology of sound patterns rather than discrete words, Voiceitt can interpret words not included within a user’s dictionary. Voiceitt is indented to be as smart as the person using the solution.

  • 3

    Versatility/accessibility

    Voiceitt can work with any mobile device. It is thus portable, marked by simplicity and ease of use with hands free functionality conducive to natural conversation. Moreover, it releases the stigma that arises from using communication devices whose conspicuous bulkiness highlights the differences between the user and its counterparts. The on-device, offline recognition, allows for nearly natural speed by reducing latency by at least 6X.

  • 4

    Enhancing sense of independence

    Voiceitt encourages people to speak with the byproduct of facilitating improvement of their speech capabilities. This restores confidence and independence, enabling users to join professional and recreational situations, to communicate with their peers, express their basic and immediate needs with ease, and improve their quality of life.

  • 5

    Security for loved ones

    Voiceitt provides security for families knowing that they have a reliable solution for their loved ones, enabling them to communicate when needed.

THE PROBLEM

Millions of children and adults find verbal communication a nearly insurmountable task. The underlying causes of their speech disorders vary widely, including degenerative diseases like Lou Gehrig’s (ALS), multiple sclerosis, and Parkinson’s; and developmental disorders like autism spectrum disorder (ASD) and cerebral palsy (CP). Voiceitt’s target market is the voice-disordered population worldwide, as well as their caregivers and families, voice researchers, speech therapists and medical professionals. The most relevant market segment for initial introduction is cerebral palsy, as these patients contain a younger group that has current access and experience using technological solutions allowing us to pilot Voiceitt with them as our initial point of market entry, with plans to enter the Parkinson’s, stroke, autism and ALS markets soon there afterwards.

LIMITATIONS OF CURRENT TECHNOLOGIES

Millions of children and adults find verbal communication a nearly insurmountable task. The underlying causes of their speech disorders vary widely, including degenerative diseases like Lou Gehrig’s (ALS), multiple sclerosis, and Parkinson’s; and developmental disorders like autism spectrum disorder (ASD) and cerebral palsy (CP). Voiceitt’s target market is the voice-disordered population worldwide, as well as their caregivers and families, voice researchers, speech therapists and medical professionals. The most relevant market segment for initial introduction is cerebral palsy, as these patients contain a younger group that has current access and experience using technological solutions allowing us to pilot Voiceitt with them as our initial point of market entry, with plans to enter the Parkinson’s, stroke, autism and ALS markets soon there afterwards.

WHAT IS VOICEITT BASIC

Voiceitt is a smart device software application that translates unintelligible sounds into clear speech in real time. Voiceitt’s first goal is to foster inclusion and allow the disabled to be independent and improve their quality of life by enabling those with motor, speech or cognitive disabilities to communicate with caregivers, family members, health care professionals and society as a whole. Voiceitt uses pattern classification technology, personalized for each individual speaker, is hands-free during use, and works in any language.

DIFFERENTIATING FEATURES

Unique features of Voiceitt representing key points of difference from industry incumbents include the following:

Personalized

The pattern classification algorithm is customized per each user’s unique speech patterns.

Language-independent

Voiceitt Basic works in any language “In-time”: Voiceitt Basic allows for instant, real-time translation of incoherent speech patterns, allowing for more fluid, natural dialogue.

Adaptive learning

Adapts to the person’s unique speech patterns in cases where the disease progresses, eliminating the need for further calibration.

Completely hands-free / voice-activated

Especially relevant, essential for users with motor control deficiencies.

Data mining

Voice data collected anonymously in Voiceitt’s servers during use serves three purposes: a) improves speech recognition capabilities as the person uses Voiceitt, b) automatic voice banking when used on “stealth mode” before voice deterioration; and c) data may be analyzed for future speech-disordered research, screening and diagnostics.

Intuitive, simple interface

Voiceitt Basic does not require any technical expertise or programming and allows for easy adoption by strangers or family members unfamiliar with AAC devices.

Compatibility

Calibration works with any smart phone or tablet; no need for additional device purchase, or expensive hardware; on-device recognition decreases latency for more spontaneous expression.

Automatic sync

Connects seamlessly to smartphones or tablets upon log-in.

VOICEITT TECHNOLOGY

The conceptual basis for Voiceitt’s algorithm lies in the observation that individuals with speech that is unintelligible to outsiders may actually be understood by one or a few of the people closest to them for example: a spouse, a mother, a caregiver. By continual exposure or conscientious attention, they’ve learned their loved one’s idiosyncratic, but recurring, speech patterns. From these observations of real-world scenarios, Voiceitt has been able to construct software that mimics this human speech-learning process via our complex algorithms which then frees the person from only communicating with their intimate circle allowing them the ability to integrate themselves into a much broader spectrum of opportunities and situations.



While considerable attention has been paid to the field of automatic speech recognition (ASR), standard ASR works only for standard speech and is therefore generally useful only to disabled individuals whose physical impairments do not extend to speech. Customization of the algorithms is vital for recognition among people with speech impairments whose speech varies widely both in types of speech impairment, or dysarthria, and in level of severity (mild, severe, and very severe). The individual’s speech may also be variable throughout the day, and commercially available ASR requires consistency of speech between and within individuals.
Our innovative approach is based on multi-domain signal processing, and pruning of dynamic voice pattern classifier search space, boosting recognition by adaptive learning of speech patterns. The user can use the device as an intermediary to provide better agency over speech capabilities in either face-to-face personal interactions or when interacting with voice-based technologies (such as: smart homes, voice controlled phone calls, web browsing and social media). Our aim is to enable better social interactions and daily independence among users.
We believe the frustration that stems from a lack of ability to communicate normally touches so many of us and it make us emotionally connected to the vision of enabling people with speech disabilities to speak freely and be understood by others around us. This vision is reinforced by the great interest and excitement of the professional community (speech therapist, occupational therapist and associations) from Israel, Europe and the US. The obvious enthusiasm and interaction with the professional community unlocked our imaginations towards creating a voice-based solution for the broader population with disabilities.

CALIBRATING TALKITT

Voiceitt Basic’s calibration stage, or initialization, is an intuitive, simple, and quick process taking up to 30 minutes for a personalized dictionary of 20-25 words and/or short phrases. The user creates a dictionary of words or short phrases –the calibration phase– which can be updated, referred to, or maintained at any given time, on an ongoing basis. Once the user’s personalized dictionary has been calibrated, the application can recall the words or phrases of the user’s individual speech and translate the unique utterances into distinguishable phrases using a voice or text output (recognition stage). For those users with severe motor or cognitive disabilities, the calibration stage will require the help of a caregiver or family member in preparing the personalized dictionary.
Consequently, the use of the app in the recognition stage can be conducted hands-free, allowing individuals with severe motor disabilities the opportunity to navigate the app independently and freely without the use of touch or fine motor skills.

Moto G4

• Works on all major carries
• Turbo charging enabled
• Fast octa-core processor
• 13 MP rear camera, 5 MP front camera
• Large 5.5 inch display
• Water repellent
• Expandable memory

LG G Pad X II

• 10.1″ tablet
• WiFi + 4G LTE
• Quad-core processor
• HD screen resolution
• 32GB internal storage, plus MicroSD slot
• 8 MP rear camera / 2 MP front camera
• T-Mobile and AT&T Compatible
• WiFi + 4G LTE