With the various options you have for hosting their AI platforms, you must ensure the method you select actually meets your needs. Today, we will look at one of these options, in particular, Edge SDKs, and outline three times this might be the right hosting method for them.
1. You want to run AI on your mobile or edge device.
As the name suggests, edge SDKs are made to allow for artificial intelligence (AI) to be deployed on edge devices, like smartphones. SDKs act as building blocks for developers, allowing them to easily integrate certain functions into their mobile apps. With Clarifai’s Edge SDKs, for instance, you can to embed computer vision into your mobile app, allowing for local, real-time image processing on any edge devices with that application.
While Edge SDKs don’t allow for custom training within the SDK itself, they can perform predictions using custom models. That is, you can build and train your own model with concepts that are relevant to your app, that can then be accessed and used by the app through the Edge SDK.
To build a custom Clarifai model, all you need to do is sign up for our Cloud API. You can use your edge device for custom training. While you’ll need to have internet access to use the API, once the custom model has been created, the SDK will be able to use it to process images with zero lag, online or offline, right in the app itself.
This brings us to our next point.
2. You and your users don’t always have a reliable internet connection.
One of the best features of edge SDKs is that they will still work even when you don’t have internet access. While internet access has boomed in the last two decades, the internet penetration rate still remains at only 56%. The penetration rate percentage is further muddled when you consider that Asia accounts for almost half that total number. In comparison, in 2017, Deloitte found that some 80% of the world’s population has a smartphone, despite such devices only being introduced in the last decade.
Edge SDKs allow developers around the world to utilize this wide dissemination of smartphones to make AI technologies like computer vision globally accessible through mobile apps. i-Nside’s diagnostic platform is a key example of this in action, allowing nurses and general practitioners to offer specialist care in even the most remote parts of the world, with their smartphone, a scope attachment, and a simple mobile app, embedded with world-class computer vision.
3. You want to keep the data private to the device.
As smartphones have become more widely available, so has photography. Thanks to these devices, 85% o the 1.2 trillion digital photos taken in 2017 were taken with smartphones. From family photos to first days in school, we now use photos to capture all sorts of special moments. That doesn’t mean, however, that we want to share them with the world. Edge SDKs allow you to protect that privacy within your apps.
When an image is processed through an Edge SDK like Clarifai’s on an edge device, that image will not leave that device, unless you explicitly allow that to happen. For instance, let’s say you built an app that could analyze any given photo, and attach a relevant quote or filter based on what it saw. While the SDK would need to analyze the image to make real-time predictions based on the trained concepts of the model in use, the processed image isn’t saved to the SDK. App developers can decide what options to offer users regarding the processed images. For instance, Clarity, our just released, open-source demo app (built for Android and iOS) analyzes the view from your smartphone camera. However, there is no option to save the image within the app itself, as for the app’s purposes, this wasn’t necessary.
Even when you connect your edge devices to the internet and our Cloud API, the images are still not saved. Instead, you’ll be able to both access and add to the aggregate learning in the Clarifai Cloud (like what every user’s app learned from the images they uploaded,) and so make your edge device a little smarter.
While Edge SDKs like Clarifai’s enable your edge devices to learn from and respond to your images both online and offline, they also offer us another way to connect with our world. With internet access not being required, the Edge SDK offers developers and users around the world access to world-leading AI technologies like Clarifai’s computer vision in the palm of their hand.