Category Archives: Speech-to-text

Another Step Forward in Automatic Speechreading

Bit by bit, computer vision analysis is advancing to the point of offering rapid, accurate speechreading. Cornell’s LipNet uses sentence-length, natural utterances to improve its accuracy, which at 93%, surpasses skilled humans. Just a prototype for now, we hope to … Continue reading

Posted in Communication, Computer vision, Deaf, Hard of hearing, Just a prototype so far, Redundant output, Speech-to-text | Leave a comment

Cloudy captions

Online captions make videos accessible to people with hearing or language disabilities, and make content indexing and searching possible for everyone.  There are more than a few online captioning tools.  But SpeakerText is a bit different.  It uses speech recognition … Continue reading

Posted in Cognitively impaired, Content, Deaf, Education, Entertainment, Hard of hearing, Information management, Network, Redundant output, Speech-to-text | Leave a comment

Tune in, turn on

Auto-Tune is software that can make bad singers sound competent, or game show hosts sound stoned. Wonder if the technology could also be used to modify the voices of people with dysarthric speech so that speech recognition applications would recognize … Continue reading

Posted in Information management, Promising component, Software, Speech impaired, Speech-to-text | Leave a comment

PhoneTag Live, you’re it!

Ditech Networks (not the lenders) has had a voice mail transcription service for a while.  Now they’re offering live transcription for conference calls.  This is a mainstream version of the real-time captioning service that’s been available for deaf and hard … Continue reading

Posted in Communication, Content, Deaf, Hard of hearing, Network, Redundant output, Speech-to-text | Leave a comment

Talk radius

Hmmm…a headset that bypasses touch screen access problems by enabling both audio output and speech input? Sounds like a product designed for blind (and some learning/cognitively disabled) folks, but it’s actually a mainstream product from Bluetrek that was shown at … Continue reading

Posted in Blind, Cognitively impaired, Hardware, Information management, Redundant input, Redundant output, Speech-to-text, Text-to-speech | Leave a comment

Thus voicemailed Zarathustra

GotVoice is the forthcoming result of a collaboration between Microsoft and a company called Spoken to create a better strategy for doing audio-to-text transcription of voicemails. Unlike some current systems that rely heavily on human transcription, GotVoice will use a … Continue reading

Posted in Cognitively impaired, Communication, Deaf, Market drivers, Redundant output, Software, Speech-to-text | Leave a comment