ibm bootcamp - text to speech api lab
TRANSCRIPT
1
Colin McCabe@beekman33IBM North America Cloud Technical Evangelist
IBM BootcampBuilding an app with a Watson API: Text to Speech Deep Dive
2
Introduction1) Today we will take an up close look at the Text-to-Speech Watson API2) Build your own Speech Synthesis App in Bluemix
3
IntroductionThe Science Behind Text to SpeechThe Text to Speech service relies on acoustic models based on decision trees. The models generate acoustic candidates to perform searches based on contextual features of the text. The system uses time-domain Pitch Synchronous Overlap and Add (PSOLA) - It can modify the pitch and duration of a speech signal. The service uses a text-processing front-end to parse the text before synthesizing it into audio form.
Based on research papers:•Database Mining for Flexible Concatenative Text-to-Speech•Automatic Exploration of Corpus-Specific Properties •Evaluation of the IBM Expressive Text-to-Speech Synthesis System •The IBM Submission to the 2008 Text-to-Speech Blizzard Challenge •The IBM expressive text-to-speech synthesis system for American English• Frequency warping based on mapping formant parameters.
4
Let’s CreateSign-up For Bluemix
API Access
Build anApp
Edit andControl Launch
1 2 3 4 5
5
Sign-up for BluemixSign-up For Bluemix
1. Bluemix.netTop Right > Sign Up
6
Dashboard – Select API2. Click Catalog at the Top
7
3. Select Text to Speech
Sign-up For Bluemix
API Access
API Access
8
4. Create The API Service (and your credentials)
95. Get your Service Credentials
106. Copy and Paste username and password (Sticky!)
11
CURL CHEATSHEEThttps://goo.gl/or7e4S
7. Test out some calls
12
HURL.IT
8. Windows users can try Hurl.it to test API
13
Starter App – GitHubSign-up For Bluemix
API Access
Build anApp
https://github.com/watson-developer-cloud/text-to-speech-nodejs
9. Google watson developer text to speech nodejs
14
Deploy to Bluemix
10. Name your app
15
Add Git11. Add GIT to prepare you DevOps tools
16
Using BluemixSign-up For Bluemix
API Access
Build anApp
Edit andControl
12. Click Edit Code
17
WEB IDE
13. Update your API credentials (app.js) and modify HTML (index.html)
18
DOMAINS / MOBILESign-up For Bluemix
API Access
Build anApp
Edit andControl Launch
14. Manage Organization / Domains15. Try it on Mobile / Download VLC to play OGG
19
Done!Congratulations, you’ve used the Watson API: Text to Speech, and built your own app.
20
Wrap-upSign-up For Bluemix
API Access
Build anApp
Edit andControl Launch
• Registration• Dashboard &
Catalog
• Select Text toSpeech API
• Create API Service
• Get ServiceCredentials
• Test API withCURL
• GitHub Starter• Name App• Add Git
(DevOps)
• Edit Code• Add API
Credentials• Manipulate
HTML
• Mobile• Domain
Name
21
Edit the manifest.yml file and change the <application-name> to something unique.applications:- services:- text-to-speech-servicename: <application-name>command: node app.jspath: .memory: 256M
The name you use will determinate your application url initially, e.g. <application-name>.mybluemix.net.Connect to Bluemix in the command line tool.$ cf api https://api.ng.bluemix.net$ cf login -u <your user ID>Create the Text to Speech service in Bluemix.$ cf create-service text_to_speech standard text-to-speech-servicePush it live!$ cf push
Command Line Option
22
Thanks
Colin McCabe@beekman33IBM North America Cloud Technical Evangelist