Node-RED Alchemy Vision image-to-speech advanced Lab

Overview

The Alchemy service has been deprecated. This lab is available for reference only. Please refer to the Visual Recognition lab for image analysis.

In this lab we will explore how Node-RED rapid prototyping capabilities allow to easily combine several Watson services into a single application using Node-RED rapid prototyping capabilities.

This lab will focus on bringing audio to the previously developed Alchemy Vision Enhanced lab (alchemy_image_analysis_enh), by reusing the service developed in the Text_to_Speech lab (lab_text_to_speech)

The Alchemy Vision API returns a wealth of information when celebrities' faces are detected on an image, including their names and related web URLs. We will use this input, presented in the identity attribute to speak up the names.

Prerequisite

The text-to-speech audio service should be deployed at the /tts/sayit URL, using the flow from the TTS lab at tts_lab_with_param_check.json

We will then modify the Alchemy Vision Enhanced lab flow to add simple audio to it, so the base of this lab will be the flow from alchvis_lab_webfaces_thumbs.json

Flow building

• Import into a new flow from the alchvis_lab_webfaces_thumbs.json flow.
• To avoid URL conflicts, rename the HTTP Input node URL's to /alchvoice, deploy and test that this works.
• To add the audio playback capability, we will simply use the HTML <audio> tag, similarly to what has been done in the web TTS lab.
• Edit the Report Faces template node. We will add a column to that table, which will conditionally add an audio tag to speak out the names of the personalities when identified by the Alchemy Vision API. We'll add a Say it! column to the table:

<th rowspan='2'>Say it</th>

and populate it with an audio tag that refers to the previously built TTS flow when the identity attribute is set:

<td><audio src="/talk/sayit?text_to_say={{identity.name}}" controls></audio></td>

Start the lab flow by pointing at ./alchvoiceURL, the output table will show additional audio controls for recognized personalities:

The completed flow is available at alchvis_image_to_speech.json