MIT 6.S191 - Automatic Speech Recognition

Explore automatic speech recognition in this MIT 6.S191 lecture featuring Rev.com experts Miguel Jetté and Jennifer Drexler. Delve into how Rev.com combines human-in-the-loop techniques with deep learning to create a cutting-edge English speech recognition engine. Learn about word error rates, data selection, speech input processing, subword units, melscale encoding, decoder mechanisms, attention-based ASR, connectionist temporal classification, and language models. Gain insights into the latest advancements in speech recognition technology and its practical applications in the industry.

Intro
Rev Data
Word Error Rate
Organization Entity
Test Benchmark
Data Selection
Speech Input
Subword Units
Melscale
Encoder Decoder
Speech Recognition
AttentionBased ASR
ConnectionistTemporal Classification
Language Models
Questions