<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Pankaj Singh on Medium]]></title>
        <description><![CDATA[Stories by Pankaj Singh on Medium]]></description>
        <link>https://medium.com/@pansingh_300?source=rss-d1b0600a5aa8------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*rjL0Olz_Vdbwfc95w2GAlQ.png</url>
            <title>Stories by Pankaj Singh on Medium</title>
            <link>https://medium.com/@pansingh_300?source=rss-d1b0600a5aa8------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 22:01:27 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@pansingh_300/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Adobe Experience Manager Assets with Microsoft Azure Face Recognition Services]]></title>
            <link>https://medium.com/adobetech/aem-assets-with-face-recognition-4cbc11596c1c?source=rss-d1b0600a5aa8------2</link>
            <guid isPermaLink="false">https://medium.com/p/4cbc11596c1c</guid>
            <category><![CDATA[azure-face-api]]></category>
            <category><![CDATA[adobe-experience-manager]]></category>
            <category><![CDATA[facial-recognition]]></category>
            <category><![CDATA[adobe-experience-cloud]]></category>
            <category><![CDATA[microsoft]]></category>
            <dc:creator><![CDATA[Pankaj Singh]]></dc:creator>
            <pubDate>Tue, 14 Mar 2023 14:58:43 GMT</pubDate>
            <atom:updated>2023-03-14T14:58:43.463Z</atom:updated>
            <content:encoded><![CDATA[<p>This article will demonstrate how to integrate Microsoft Azure Face Recognition services with AEM Assets to smartly identify faces in uploaded images and take appropriate action. While AEM Assets already includes object recognition through Adobe Sensei services, the focus here will be on identifying specific people/faces in the image using Microsoft Azure Face Recognition Services. Upon successful identification, the image will be moved to a restricted folder within AEM Asset.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZYgfWh3aNxMY3bgupTCsig.png" /></figure><h3>Technical Details</h3><h3>You’ll need:</h3><ul><li>Adobe Experience Manager</li><li>Adobe Microsoft Azure account with Face Recognition Service</li></ul><h3>Steps towards Solution</h3><p>We can divide the entire implementation into four major steps :</p><ol><li><strong>Enable Face Recognition Service in Microsoft Azure Account.</strong></li><li><strong>Create custom fields in Folder Metadata in AEM and apply it on specific folders.</strong></li><li><strong>Create Workflow in AEM to train the Face Recognition Model.</strong></li><li><strong>Add Custom Workflow step in AEM.</strong></li><li><strong>Trigger Training Workflow on the Training Folder.</strong></li><li><strong>Create Custom Workflow Step to detect Faces in the image.</strong></li><li><strong>Use the detection process step in workflow.</strong></li><li><strong>Expose configuration values via OSGi (optional).</strong></li></ol><h3>Implementation</h3><p><strong>Enable Face Recognition Service in Microsoft Azure Account</strong></p><p>For this use case, you will utilize Microsoft’s “Cognitive Services — Face” among several services offered in Microsoft Azure. You will need to activate this service on your account. To create a new instance, select a Resource Group and provide an instance name.</p><p>The instance name you provide will be used to construct the API URL that will grant access to the service.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZhRgfo3fJaqNKAT2192U9A.png" /><figcaption>Create Face Service in Microsoft Azure Account</figcaption></figure><p>In order to access the face service, it is necessary to include an authentication key, known as “Ocp-Apim-Subscription-Key”, in the header of each API call made to the service.</p><p>This key can be obtained from the “Keys and Endpoint” section of the newly created face service. Simply copy the highlighted key (shown in red) and save it for later use.</p><p>To access the face service, an authentication key known as “Ocp-Apim-Subscription-Key” must be included in the header of each API call. This key can be obtained from the “Keys and Endpoint” section of the newly created face service by copying the highlighted key (shown in red) and saving it for later use.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DgRNpol_pgeeG6iP4UFhiQ.png" /></figure><p><strong>Create custom fields in Folder Metadata in AEM and apply it on specific folders</strong></p><p>There are different ways to provide input for the Face Service Model API, but we will leverage the flexibility of customizing the Folder Metadata Schema in AEM. You need to create a custom schema for folder metadata to include fields for <em>Person Name</em> and <em>Person Description</em>. Subsequently, we’ll apply this schema to the folder in AEM assets that we intend to use for training the Face recognition model.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PJCda4v_Wwka53CRI_PmoQ.png" /><figcaption>Custom Folder Metadata Schema with Text Fields</figcaption></figure><p>Within the “Face Recognition Training” folder, you can create a subfolder and upload a minimum of five images that will be used to train the model. It is important to ensure that the uploaded images are clear and belong to the same person/face, as these images will be utilized for the training process.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Znh4rpbjhPEhN3TqQwm6Ew.png" /></figure><p>Edit your folder properties where we can provide details in the field which were added using Folder Metadata Schema.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gLJyQjyC9iWBjR9p814oNg.png" /></figure><p><strong>Create workflow in AEM to train the Face Recognition Model</strong></p><p>To train the system, you will need to write a custom process step that utilizes Microsoft Face API calls. This process step will be triggered when the workflow is initiated and will be responsible for reading all images in the designated folder. The APIs will need to be called in a specific order, as follows :</p><ol><li>Use the Microsoft Cognitive Services API to create a group — <a href="https://learn.microsoft.com/en-us/rest/api/faceapi/person-group/create?tabs=HTTP">Microsoft Cognitive Services — Create Person Group</a>.</li><li>Use the Microsoft Cognitive Services API to create a person — <a href="https://learn.microsoft.com/en-us/rest/api/faceapi/person-group-person/create?tabs=HTTP">Microsoft Cognitive Services — Create Person</a></li><li>Use the Microsoft Cognitive Services API to add the person’s face — <a href="https://learn.microsoft.com/en-us/rest/api/faceapi/person-group-person/add-face-from-stream?tabs=HTTP">Microsoft Cognitive Services — Add Face From Stream</a></li><li>Use the Microsoft Cognitive Services API to train the group model — <a href="https://learn.microsoft.com/en-us/rest/api/faceapi/person-group/train?tabs=HTTP">Microsoft Cognitive Services — Person Group Train</a></li></ol><p>Sample Code for entire flow which needs to go as Custom Workflow process step :</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/eaf44805b8b88a89fdc80f7b55b7e919/href">https://medium.com/media/eaf44805b8b88a89fdc80f7b55b7e919/href</a></iframe><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/73d4f1ff3055f429c50b77e6039be742/href">https://medium.com/media/73d4f1ff3055f429c50b77e6039be742/href</a></iframe><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/8bec2abf84d5e53310cb1a882e1e9278/href">https://medium.com/media/8bec2abf84d5e53310cb1a882e1e9278/href</a></iframe><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/fe76812ea4ddae14b831c08154758df4/href">https://medium.com/media/fe76812ea4ddae14b831c08154758df4/href</a></iframe><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/e5c3d87f2c8e5c349b4308830e313784/href">https://medium.com/media/e5c3d87f2c8e5c349b4308830e313784/href</a></iframe><p><strong>Add custom workflow step in AEM</strong></p><p>Within AEM, design a workflow model that includes your custom process as the selected process step.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-5KNu7beztbW1oZlzZWxhw.jpeg" /></figure><p><strong>Trigger Training Workflow on the Training Folder</strong></p><p>Choose the folder that was created for training and initiate the “Face Recognition Training Workflow” on that particular folder.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yiiUICbYKrC7U3DgD_lXqA.png" /></figure><p>Once the workflow is completed, you will receive a notification within AEM indicating whether the workflow has finished successfully or has failed.</p><p><strong>Create Custom Workflow Step to detect Faces in the image</strong></p><p>Now that you are familiar with how to create a custom process step, you can use it to make the following two MS Face API calls for detecting/identifying faces from the provided image:</p><ol><li>Use the Microsoft Cognitive Services API to detect a face - <a href="https://learn.microsoft.com/en-us/rest/api/faceapi/face/detect-with-stream?tabs=HTTP">Microsoft Cognitive Services — Detect Face</a></li></ol><p>You will need to pass the image’s stream to this API, and it will return an array of faces detected in the image.</p><p>2. Use the Microsoft Cognitive Services API to identify a face - <a href="https://learn.microsoft.com/en-us/rest/api/faceapi/face/identify?tabs=HTTP">Microsoft Cognitive Services — Face Identify</a></p><p>You will need to provide an array of <em>faceIds</em> from the result of the previous call, along with <em>personGroupId</em>, <em>maxNumOfCandidatesReturned</em>, and <em>confidenceThreshold</em>. This API will return an <em>array of FaceIds</em> and associated <em>PersonIds</em>, along with a confidence score. If the confidence score meets your threshold, it indicates that the detected face matches the face your model was trained with.</p><p><strong>Use the detection process step in workflow</strong></p><p>You can utilize the custom process step that you have created by either integrating it into a custom workflow model or updating the “Dam Update Asset” workflow depending on how you would want to leverage the detection.</p><p>Once the matching face (based on the previous training) is detected, it is up to you to decide what action to take. You may choose to move the uploaded image to a different folder, notify the author that the image has an issue, or subject the uploaded image to additional approvals, among other possible actions.</p><p><strong>Expose configuration values via OSGi (optional)</strong></p><p>As a recommended practice, you may want to expose all the configurable values needed in the API calls as OSGi configurations which makes it easy to change the values at runtime.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*c00qNhiF7GAk923lL0ZEXQ.png" /></figure><p><em>Disclaimer : For production usages, one should consider additional aspects e.g. security, performance etc.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4cbc11596c1c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/adobetech/aem-assets-with-face-recognition-4cbc11596c1c">Adobe Experience Manager Assets with Microsoft Azure Face Recognition Services</a> was originally published in <a href="https://medium.com/adobetech">Adobe Tech Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AEM Forms with Adobe I/O Runtime]]></title>
            <link>https://medium.com/adobetech/aem-forms-with-adobe-i-o-runtime-61e0e039e036?source=rss-d1b0600a5aa8------2</link>
            <guid isPermaLink="false">https://medium.com/p/61e0e039e036</guid>
            <category><![CDATA[adobe-io]]></category>
            <category><![CDATA[aem]]></category>
            <category><![CDATA[aem-forms]]></category>
            <category><![CDATA[adobe-experience-cloud]]></category>
            <category><![CDATA[serverless]]></category>
            <dc:creator><![CDATA[Pankaj Singh]]></dc:creator>
            <pubDate>Mon, 07 Sep 2020 17:38:57 GMT</pubDate>
            <atom:updated>2020-09-07T17:38:57.814Z</atom:updated>
            <content:encoded><![CDATA[<p>We often come across use cases around pre-filling and the submission of data using AEM Forms from/to external databases (DBs). Out of many possible solutions possible to address such use-cases, one solid contender is a modular services based approach using AEM Forms in conjunction with Adobe I/O Runtime.</p><p>Adobe I/O Runtime is a serverless architecture with the benefits of flexibility, scalability, reduced operating cost, etc. Modular services deployed on Adobe I/O Runtime communicates with an external DB and processes the data retrieved from DB thereafter.</p><p>Meanwhile, AEM Forms contains powerful authoring capabilities. It has an intuitive UI that saves time in designing and publishing the form by reducing custom development effort.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fZaGutHZLKZ6DhExI5YmAw.jpeg" /></figure><h3>Technical Details</h3><h3>To use this solution, you’ll need:</h3><ul><li>Adobe Experience Manager Forms</li><li>Adobe IO Runtime</li><li>Any RDBMS (e.g. MySQL)</li></ul><h3>Steps to implmenting the solution</h3><p>We can follow a step-by-step approach to connect each solution with the others.</p><ol><li><strong>Deploy Service at Adobe Runtime I/O </strong><em>(Creation of a service that will connect with MySql database and will be exposed as REST).</em></li><li><strong>Create a form with AEM </strong><em>(Author a form with AEM </em>Forms<em> that can use some rules for prefilling the form on load, which will be submitted using Adobe I/O Runtime).</em></li><li><strong>Set up a DB hosted externally.</strong></li></ol><h3>Implementation</h3><h4><strong>Deploy Service at Adobe Runtime I/O</strong></h4><p>We will create two services to deploy on Adobe I/O Runtime.</p><blockquote><em>Note: For simplicity, we are creating two separate services; however, we can combine both into one service as well.</em></blockquote><p>Create <em>db-insert.js</em> file with below code (update the DB details and query)</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/db0905cc34f0ee51bfc27f7003c4c5ba/href">https://medium.com/media/db0905cc34f0ee51bfc27f7003c4c5ba/href</a></iframe><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d48f63351c3c9df7ac6a43b485bbec63/href">https://medium.com/media/d48f63351c3c9df7ac6a43b485bbec63/href</a></iframe><p>js files can be deployed both as runtime actions and exposed as REST endpoints using OpenWhisk CLI.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/efb0b7cd132215abe87b9796a839e2c8/href">https://medium.com/media/efb0b7cd132215abe87b9796a839e2c8/href</a></iframe><p>The “REST endpoint URL” mentioned in line# 8 in the snippet above is the URL that we will use to insert the form data (first snippet) and retrieve it from DB (in the second snippet).</p><h4><strong>Create a form with AEM</strong></h4><p>Create a simple form that takes user details input.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*u3KbSGO4C8cFCzxblRX-vw.png" /></figure><p>There are two ways to post to the Adobe I/O Runtime service from this form. Both ways work effectively the same and selecting anyone from these depends on the flexibility we need with the submission(in terms of data manipulation before/after the form submission etc.).</p><p>a) Write a custom rule on click of submit button (e.g. snippet below) :</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/6375512bcfbafc0aa548235a55812d57/href">https://medium.com/media/6375512bcfbafc0aa548235a55812d57/href</a></iframe><p>b) Alternatively, configuring “Submit to REST endpoint” in submit action of the form (refer screenshot below):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XZGbUewg0OnmX514rcX_cw.png" /></figure><p>Now, within the same form, we can create a rule on the root panel to fetch user details from DB via Adobe I/O Runtime and populate into the form fields. Following code snippet can be used on initialization of the root panel of form which then prefills the first name, last name, and city fields by using the logged-in user’s id :</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/7ea4e2b71419569785d36b355395479b/href">https://medium.com/media/7ea4e2b71419569785d36b355395479b/href</a></iframe><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1iweEn0r42B-d5zIVvKVsA.png" /></figure><h3>What’s Next :</h3><p>Similar to the approach discussed above, we have endless possibilities to integrate AEM with other Experience Cloud Products (Adobe Target, Campaign, etc.) leveraging the Adobe I/O Runtime.</p><p><em>Disclaimer : For production usages, one should consider additional aspects e.g. securing IO Runtime actions, performance metrics among others.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=61e0e039e036" width="1" height="1" alt=""><hr><p><a href="https://medium.com/adobetech/aem-forms-with-adobe-i-o-runtime-61e0e039e036">AEM Forms with Adobe I/O Runtime</a> was originally published in <a href="https://medium.com/adobetech">Adobe Tech Blog</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>