Using Automated Speech Recognition to Support Human Transcriptions

In early 2020 Aigent built a Speech Model with a Word Error Rate (WER) of 9%. For those new to Speech Recognition, WER is a common metric used to measure the accuracy of speech model transcriptions. Aigent’s Speech Model’s WER is particularly impressive given the model is trained upon data from a notoriously tricky domain – …

Using Automated Speech Recognition to Support Human Transcriptions Read More »