In plain language:
When a person speaks, the message carries emotional and prosodic information alongside its words. Conventional machine translation, in its early years, captured the words. The patented method combined sentiment analysis with machine translation to capture the emotional content during translation: representing it as structured data bound to the specific words and phrases that produced it, and carrying it across the language boundary with the message, so that a recipient receives not only what was said, but the manner in which it was said.
The specification further contemplates representations that carry adjusted vocal delivery in the target language, and from that, a holographic presentation of the speaker.
Licensing, sublicensing and research collaboration are directed to the founder and inventor.
