posted on 2018-07-16, 07:55authored byRafael Dantas, Chris Exton, Andrew Le Gear
Even with advances in modern GSM technologies there remains situations where increased data compression rates can be valuable (Dantas, Exton, and Le Gear, 2018). While the adoption of 4G networks is increasing, there is still a significant portion of the world’s mobile devices connected through less capable 2G or 3G networks.
We propose that with recent developments in speech-to-text technology, devices are now capable of correctly transcribing speech samples with enough precision to be used for telecommunications. Our initial experiments demonstrate that it is possible to achieve significantly greater compression rates by converting speech into text streams when compared to current voice codecs. Such text streams could also be further compressed using traditional text compression algorithms like Huffman encoding (Huffman, 1952).
We also propose that Blockchain technology (Nakamoto, 2008) could bring authentication functionality to a text-to-speech compressed VoIP call that would allay such concerns.
Funding
Study on Aerodynamic Characteristics Control of Slender Body Using Active Flow Control Technique