top | item 41980246

(no title)

_gmax0 | 1 year ago

Off topic, but what's the state-of-art behind speech recognition models at the moment?

Are people still using with DTW + HMMs?

discuss

order

IshKebab|1 year ago

HMMs haven't been state of the art in speech recognition for decades (I.e. since it actually got good). It's all end-to-end DNNs now. Basically raw input -> DNN -> ASCII.

Well almost anyway - last I checked they feed a Mel spectrogram into the model rather than raw audio samples.

pcwelder|1 year ago

> state of the art in speech recognition for decades

Decades doesn't sound right. Around 2019, the Jasper model was SOTA among e2e models but was still slightly behind a non e2e model with an HMM component https://arxiv.org/pdf/1904.03288