Connectionist Temporal Classification for End-to-End Speech Recognition thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Connectionist Temporal Classification for End-to-End Speech Recognition

Published on Jul 31, 20161985 Views

The performance of automatic speech recognition (ASR) has improved tremendously due to the application of deep neural networks (DNNs). Despite this progress, building a new ASR system remains a chal

Related categories

Chapter list

Connectionist Temporal Classification for End-to-End Speech Recognition00:00
Fundamental Equation of Speech Recognition00:09
Recognition Conceptually: AM and LM01:13
Hidden Markov Models02:25
Context-Dependent States02:39
Duration Modeling03:35
State-of-the-Art ASR04:51
Let’s Take a Step Back05:37
Connectionist Temporal Classification06:34
Observations07:52
Problem with Best Path Decoding08:36
Enter WFST Decoding09:34
Results on Read Speech10:48
Results on Conversational Speech12:42
CTC Conclusions14:15
Thank You!16:13