Abstract
<jats:title>Abstract</jats:title> <jats:p> This entry reviews advances in speech‐to‐text (i.e., speech recognition) and text‐to‐speech (i.e., synthesis) technologies, and how these advances may be used to develop two distinct approaches for pronunciation feedback: <jats:italic>explicit</jats:italic> feedback that uses speech <jats:italic>recognition</jats:italic> techniques to help L2 learners detect, identify, and correct pronunciation errors in their speech, and <jats:italic>implicit</jats:italic> feedback that uses speech <jats:italic>synthesis</jats:italic> techniques to generate synthetic voices that L2 learners can use as personalized models. We will provide a brief history of speech‐to‐text recognition and text‐to‐speech synthesis through the lens of computer‐assisted pronunciation training, and present two state‐of‐the‐art models based on modern deep‐learning techniques. </jats:p>