Slides from Tuesday's talk online, next talk Tara Sainath on September 16th
7 views
Skip to first unread message
Colin Raffel
Sep 10, 2015, 3:14:53 PM9/10/15
to cu-neu...@googlegroups.com
Hi all, slides from Michael Mandel's talk on Tuesday are now online here. Our next session will be next Wednesday, September 16th (less than a week away!). We're hosting Tara Sainath from Google in Costa Commons (aka CEPSR 750) at 4pm. Talk details follow, please let anyone who might be interested know. See you then!
"Towards End-To-End Speech Recognition"
Tara Sanaith, Google Research
4pm, September 16th
"Towards End-To-End Speech Recognition"
Tara Sanaith, Google Research
4pm, September 16th
Costa Commons/CEPSR 750
Abstract:
In this talk, I will discuss various efforts in our group at Google towards replacing various parts of the acoustic modeling pipeline with neural networks. First, I will describe a new modeling approach known as Convolutional, Long Short-Term Memory, Deep Neural Networks (CLDNNs), and why this architecture makes sense for speech tasks. Next, I will talk about using CLDNNs for raw-waveform modeling, allowing us to remove front-end log-mel filterbank feature computation. Finally, I will discuss CTC, which allows us to remove the need for a prior alignment and CD states.
∿
Abstract:
In this talk, I will discuss various efforts in our group at Google towards replacing various parts of the acoustic modeling pipeline with neural networks. First, I will describe a new modeling approach known as Convolutional, Long Short-Term Memory, Deep Neural Networks (CLDNNs), and why this architecture makes sense for speech tasks. Next, I will talk about using CLDNNs for raw-waveform modeling, allowing us to remove front-end log-mel filterbank feature computation. Finally, I will discuss CTC, which allows us to remove the need for a prior alignment and CD states.
∿
Reply all
Reply to author
Forward
0 new messages