Speechmatics released a new API including an
enhanced transcription engine (2h free per month!) that we integrated into the Auphonic Web Service now.
In this blog post, we also compare the accuracy of all our integrated speech recognition services and present our results.
Automatic speech recognition is most useful to make audio searchable: Even if automatically generated transcripts are not perfect and might be difficult to read (spoken text is very different from written text), they are very valuable if you try to find a specific topic within a one-hour audio file or if you need the exact ...