User Tools

Site Tools


private:low_latency_asr_for_the_deaf

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

private:low_latency_asr_for_the_deaf [2018/07/05 17:12] (current)
Line 1: Line 1:
 +===== Low latency ASR for the deaf =====
 +
 +This has roots in "the colour of speech"​ and en-to-end models on characters or phonetic spellings.
 +
 +People process words. ​ They are very used to reading and for clean input we can do low latency ASR that is only a few words behind.
 +
 +DECISION: ​ Abandon "the colour of speech"​ and en-to-end models. ​  These won't have good word breaks and so will take a long time to learn. ​  They will be overtaken by standard ASR.
 +
 +DECISION: ​ Don't do standard ASR as it competes with Speechmatics.
  
private/low_latency_asr_for_the_deaf.txt ยท Last modified: 2018/07/05 17:12 (external edit)