All of Moshi’s products harness the most advanced voice recognition technology available. The beginning of speech recognition development traces back to Alexander Grahm Bell in the 1870s. However, voice recognition technology didn’t become functional until the 1980s and useful until the 1990s.
Today, Moshi’s Technology that Listens™ or Voice Control (VC) Technology utilizes a unique blend of two separate models, which in comparison differ from any other voice recognition application. Moshi’s combination of Neural Net and Hidden Markov Modeling (HMM) Phonetic Speaker Independent (SI) Recognition Software allows anyone to be able to control their electronics using their voice alone.
The HMM recognition is a statistical model that requires zero training, zero setup, and is incredibly simple and easy to use. The combination of the HMM and Neural Net model is supported and trained to recognize over 1,000 different recorded voices. Moshi spent 3 years traveling to each region of the United States to record every signle different voice type. This allows all of Moshi’s products to be able to understand a large vocabulary of voice recognition commands. The technology is speaker independent (and not dependent) making all Moshi’s products capable of understanding almost anyone right out of the box. Moshi’s technology even accommodates a wide variety of accents too.
Overall the best part about Moshi’s products is that without any pre-programming Moshi’s technology will work with your parents, grandparents, kids, cousins, aunts, family, neighbors, and even your favorite in-laws.