In my previous blog post, you learned how to use Einstein Vision with Force.com. There are many possible use cases, but the one that we focus on today involves predicting images on Heroku, using legacy on-premises or native mobile applications. Let’s dive into some examples.
Introducing Einstein Vision wrappers for Java and Swift
You can find two different repos for Java and Swift on GitHub—each of them contain a dedicated wrapper for Einstein Vision. The wrappers are also available as compiled versions (Bintray for Java and Cocoapods for Swift), so you can easily incorporate them into any project.
“Why should I use a wrapper?” Because it makes it easier for you to get started with Einstein Vision. Consider these use cases:
- A customer hosts images within an on-premise Java application and wants to train Einstein Vision using these images from the existing Java application.
- Mobile app users can predict an image from a native mobile app (Android or iOS) without taking a round-trip to another server.
- An end-user–facing application on Heroku should be optimized to provide predictions on incoming images for a better app experience.
Check out the following videos to learn more.
Native app with Swift
This video showcases how to use the Swift wrapper within a native iOS application. With just a few lines of code, end users can quickly start predicting images.
If you want to start incorporating Einstein Vision into your Java applications, then check out this video to see how a simple Spring Boot and Vaadin application gets enhanced with image recognition capabilities.
Test it out for yourself with this live application, available on Heroku.
Moving forward from this point
These wrappers are not an official Salesforce project, but feel free to contribute on GitHub (Java/Swift), or comment on this blog post. Take a look at the new Einstein Vision add-on for Heroku, which simplifies the setup process. It is also used in the above mentioned Java web app (see on GitHub).
If you haven’t already, check out the Einstein Vision documentation or read through these awesome trails. If you want to learn more about how to use Einstein Vision with node.js, watch this webinar Build Smarter Apps with the New Predictive Vision Service. For questions around the Einstein Vision feel free to checkout the developer forum or Stack Exchange.
About the author
René Winkelmeyer is as a senior developer evangelist at Salesforce. He focuses on enterprise integrations, mobile, and security with the Salesforce Platform. You can follow him on Twitter @muenzpraeger.