2529742 / personage

VIE widget

personage

Personage is a VIE widget that allows to annotate depicted persons with schema.org structures.

It creates a VIE entity for the entire image; and also creates separate VIE entitie(s) for fragment(s) of the image corresponding to the person(s) detected. Both image- and fragment-related entities are of the http://schema.org/ImageObject VIE type.

The widget also includes a functionality for tagging the detected fragments using existing semantic markup on the page: e.g. the page contains the text with a hidden markup

<span about="http://dbpedia.org/resource/John_Lennon" typeof="Person"> John Lennon </span>

you can drag this text ("John Lennon") and drop it on the appropriate detected fragment of the image. This creates a semantic relation between two entities: person and image objects.

Another functionality allows to highlight the text and and the image, which are semanticaly related (like described above): every time the cursor hovers over the text or the fragment - both get highlighted.

The demo utilizes skybiometry: a service for face detection and face recogntion.

Example

http://2529742.github.io/personage/

1. Load the dependency scripts in the following order:

    <!-- jQuery and jQuery UI -->
    <script type="text/javascript" src="https://ajax.googleapis.com/ajax/libs/jquery/1.7.1/jquery.min.js"></script>
    <script type="text/javascript" src="https://ajax.googleapis.com/ajax/libs/jqueryui/1.8.18/jquery-ui.js"></script>
    <!-- VIE dependencies: uderderscore, backbone, rdfquery; and wigdet-specific script: livequery -->
    <script type="text/javascript" src="underscore.js"></script>	
    <script type="text/javascript" src="backbone.js"></script>
    <script type="text/javascript" src="jquery.rdfquery.js"></script>
    <script type="text/javascript" src="jquery.livequery.js"></script>
    <!-- VIE: VIE library and the widget -->
    <script type="text/javascript" src="vie-2.1.0.js"></script>
    <script type="text/javascript" src="vie.widget.personage.js"></script>
   

and create a new script for your code:

	<script type="text/javascript">
	//...your code...
	</script>

2. Write there a function that would call the widget:

    function startAnnotation(){
      	//initialize face tagger 
      	$('#photos').viePersonage({ //call the widget on the container with the images or on a particular image
      		services: { //face detection services
    				skybiometry: {
    					use: true,	// flag to indicate that this service will be used for face detection (if there are many)
    					api_key: "YOUR_API_KEY",
    					api_secret: "YOUR_API_SECRET"
    				}
      		},
      		myVIE: myVIE, //your VIE instance
      		done: imageResults, // callback function to process the resulted image fragment entities
      		//drag-and-drop and hightlight functionality for existing semantic markup on the page:
      		draggable: none, // a filter for activating draggables, e.g. for all markup with persons use draggable: [Person]
      		highlight: true // use/not use highlighting
      	});
    }
   

3. Write another function to process the results:

    function imageResults(entities){
      	for(var f = 0; f < entities.length; f++){
      		var fragment = entities[f];
    		console.log(fragment);
    	}
    }
   

4. Call the startAnnotation function whenever you need, e.g. on the button click:

    <button id="start" onclick="startAnnotation()">Start image annotation</button>