Jurassic Park Extrapolation Increases Speech to Speech Engine Accuracy

Citation: Wahlstedt, R (2014/12/01) Jurassic Park Extrapolation Increases Speech to Speech Engine Accuracy.
DOI (original publisher): 10.6084
Semantic Scholar (metadata): 10.6084
Sci-Hub (fulltext): 10.6084
Internet Archive Scholar (search for fulltext): Jurassic Park Extrapolation Increases Speech to Speech Engine Accuracy
Download: http://dx.doi.org/10.6084/m9.figshare.1275341
Tagged: Computer Science (RSS) artificial neural networks (RSS)

Summary

This paper discusses on how a person can concentrate on so much information per second. The processes of conceptualizing and formatting into words are very challenging for some people. This is what computer scientists commonly refer to as a bitrate. This paper divides technologies for two groups, those who have had trouble since birth. These people use strategies to avoid certain words, They repeat what they are saying or they rephrase it to say it a couple different ways, or they avoid pronouns. this module implements the work discussed in the book "The Secret Life of Pronouns". For people who encounter speaking difficulties later in life due to als or a stroke, the program works harder using other modules such as defining what is not the topic.

Some points are

- Use of context to predict what someone with speech impairment is or wants to say or write

- Use of computer learning algorithm to predict conversation : speech trajectory and voice shaping

- Use of Meta data for better speech prediction : grouping , chunking.

- Neuroscience of language synchronous firing chain.

- Use of Game theory for language comprehension and eliminating what is not the topic

- use of artificial neural network.