Fact · Dictation
3.7percent
In the same study, close-talking read speech recorded in a studio had a word error rate of 3.7 percent with DNN models.
Checked against its source on October 16, 2026
In the source's own words
“The Word Error Rate (WER%) obtained by decoding the close-talking WSJ sentences recorded in the recording studio is 3.7”
Realistic multi-microphone data simulation for distant speech recognition, arXiv preprint, 2017, Section 4.1.
Open the original- What it measures
- This is the baseline for the distant-microphone comparison.
- What it does not show
- Studio recording and read sentences. It is a best case, not a typical dictation result.