We make the world's literature accessible for every person (regardless of missing language skills and other perceptual abilities).
We humans live our lives speaking diverse natural and artificial languages, and we have diverse physical abilities and psychological attributes. If only we could speak and understand all these languages, have photographic memory, better attention span, perfect musical pitch, keen eyesight and so on ...
Emerging artificial intelligence technology enables us to build perceptual agents, assistants, and tools to furnish us with such human and even super-human abilities.
We train a team of specialist AI agents that co-operate in real-time to provide you with access to information in your own language, adapted to your specific physical and psychological abilities.
The perceiver agent builds a multi-facet, multi-layer representation of information content, its source, intent and context. Since we humans conceptualize our world (and imagined worlds) in terms of stories, the perceiver agent represents the information it ingests as story graphs.
The translator focuses in on each perceived information element in its context. This enables it to faithfully translate elements from the original (canonical) language to another (target) language.
The modal adapter focuses on perceived and translated information elements in their relevant context. This enables the agent to faithfully interpret and transform each element expressed in the original modality (text, audio, image) to specified target modalities.
Like the storytellers and minstrels of the past, the presenter agent interacts with you, to perform the transformed content, answer your questions and consider your feedback.
Other team members include an executive, a producer, and a researcher, but perhaps the above notable mentions suffice here.
Our current focus is to develop a non-commercial accessibility service for vision impaired individuals. We provide a blind or low vision person with the ability to access, enjoy and explore fiction and non-fiction books using conversational interaction in any language with cross-modal adaptation of text and image content into audio presentation and description.
Please address your queries and comments to info@crossmodal.io.