Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

update ctc_beam_search_decoder design doc #2423

Merged
merged 5 commits into from
Jan 31, 2018
Merged

update ctc_beam_search_decoder design doc #2423

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
c8495f8
update ctc_beam_search_decoder design doc
Jun 8, 2017
91d7838
Merge branch 'develop' of upstream into add_ctc_decoder_design_doc
Jan 31, 2018
7c53d72
Refine the design doc for ctc_beam_search_decoder
Jan 31, 2018
535f6bd
Rename the DS2' design doc
Jan 31, 2018
de1a701
Adjust the width of figure 2
Jan 31, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 12 additions & 1 deletion doc/design/speech/README.MD
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -140,7 +140,17 @@ TODO by Assignees

### Beam Search with CTC and LM

TODO by Assignees
<div align="center">
<img src="image/beam_search.png" width=400><br/>
Figure 2. Algorithm for Beam Search Decoder.
</div>

- The **Beam Search Decoder** for DS2 CTC-trained network follows the similar approach in \[[3](#references)\] with a modification for the ambiguous part, as shown in Figure 2.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What is the ambiguous part and what is the modification? Could you please add more details?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done

- An **external defined scorer** would be passed into the decoder to evaluate a candidate prefix during decoding whenever a space character appended.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

remove "defined"

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done

- Such scorer is a unified class, may consisting of language model, word count or any customed evaluators.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

  1. --> any other custom scorers.
  2. What do you mean by "a unified class"?
  3. Remove "may".

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done and modified.

- The **language model** is built from Task 5, with a parameter should be carefully tuned to achieve minimum WER/CER (c.f. Task 7)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

a parameters --> parameters

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done

- This decoder needs to perform with **high efficiency** for the convenience of parameters tuning and speech recognition in reality.


## Future Work

Expand All @@ -153,3 +163,4 @@ TODO by Assignees

1. Dario Amodei, etc., [Deep Speech 2 : End-to-End Speech Recognition in English and Mandarin](http://proceedings.mlr.press/v48/amodei16.pdf). ICML 2016.
2. Dario Amodei, etc., [Deep Speech 2 : End-to-End Speech Recognition in English and Mandarin](https://arxiv.org/abs/1512.02595). arXiv:1512.02595.
3. Awni Y. Hannun, etc. [First-Pass Large Vocabulary Continuous Speech Recognition using Bi-Directional Recurrent DNNs](https://arxiv.org/abs/1408.2873). arXiv:1408.2873
Binary file added doc/design/speech/image/beam_search.png
View file
Open in desktop
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.