Watson IBM Bluemix

Overview

Our last series of IBM Watson challenge were so successfully that we thought we'd run some more!. Again, this is a sweepstakes challenges so each week we'll randomly pick a passing submitter and give them $250. For this challenge we'll be using Node-RED along with some Watson services. Node-RED is a tool for wiring together hardware devices, APIs and online services in new and interesting ways. It's very similar in concept to Yahoo! Pipes.

The task of this challenge is to assemble 3 Watson services with Node-RED to quickly to create a machine translation from English speech to Spanish and read it out. You can either generate the speech from microphone or use preloaded audio. Here's a working example of the application - https://wb-watson-demo.mybluemix.net/

Documentation on the services:

http://www.ibm.com/smarterplanet/us/en/ibmwatson/developercloud/speech-to-text.html
http://www.ibm.com/smarterplanet/us/en/ibmwatson/developercloud/machine-translation.html
http://www.ibm.com/smarterplanet/us/en/ibmwatson/developercloud/text-to-speech.html

Steps:

--- Get the demo code for Speech to Text and follow instructions to deploy it to your Bluemix account. You can use the latest code from Github but if you want to come up with the same UX as the demo you would use an older revision https://github.com/watson-developer-cloud/speech-to-text-nodejs/tree/e9bafd71373c4157cc75dbfa6ef27a6f9636c6d5. Either version is fine as long as you accomplish the task.

--- Check out Node RED - http://nodered.org/.  Start the Node-RED Starter template from your Bluemix account and play with it.  Build a flow that receives a piece of English text from an HTTP request, routes it through Machine Translation to translate it into Spanish and then invokes Text to Speech and produce an audio file as the HTTP response.  Call your flow from browser to test.

--- Modify the Speech to Text demo code to add an html5 audio element and call your Node RED flow after the text is acquired from Speech to Text service, load the response audio file into the audio element and auto play it. It should be just couple of lines of code and that’s all!

Help Documentation