flâneur — a map of the web's best reading

Sequence Modeling with CTC

distill.pub · 7,777 words · saved by 2 readers

A visual guide to Connectionist Temporal Classification, an algorithm used to train deep neural networks in speech recognition, handwriting recognition and other sequence problems.

Sequence Modeling with CTC Sequence Modeling With CTC A visual guide to Connectionist Temporal Classification, an algorithm used to train deep neural networks in speech recognition, handwriting recognition and other sequence problems. How CTC collapsing works For an input, like speech Predict a sequence of tokens Use return to input a blank ( ϵ ) (\epsilon) ( ϵ ) Merge repeats, drop ϵ \epsilon ϵ Final output Authors Affiliations Awni Hannun Stanford University Published Nov. 27, 2017 DOI 10.23915/distill.00008 Introduction Consider speech recognition. We have a dataset of audio clips and corre

Explore this link on the map →

saved by

related reading