Augmented Reality and Machine Learning are two disruptive technologies that are shaping new digital operating models. Augmented Reality bridges the gap between the physical and virtual world by overlaying digital snippets above physical objects. Apple’s CEO Tim Cook regards Augmented Reality as a big idea like the smartphone.
Machine learning, on the other hand, picks up patterns from data to arrive at predictions. Access to large data lakes, easiness of building data pipelines, increase in real-time data processing capabilities, all would make ML a powerful mainstream technology. It is no surprise that Sundar Pichai has put Google on “AI-first” footing.
Putting AR and ML together, we get a dynamic combo that could be a solution for many pressing problems that diverse industries face.
The visual data that AR apps/programs collect over a period can be used to train data models for image sensing. Such a trained model can be of immense benefit to industries like manufacturing, retail, construction, education, transportation and much more.
Goldman Sachs estimates that immersive technologies like AR and VR will grow into an $80 billion market by 2025, or “about the size of the desktop market as of today.” Studies were done by World Economic Forum along with Goldman Sachs also found that these immersive technologies are enjoying strong demands from multiple creative industries including entertainment (video games), live events, video entertainment and retail among many others.
There is also a strong public opinion that VR & AR are meant for gaming and entertainment. But, these technologies do have the power to have a severe impact on heavy-weight and process-centric industries like real-estate, marketing, manufacturing, healthcare, defense, etc.
Augmented reality is at the tipping point of becoming a mainstream technology. The latest iPhone X with AR is just the beginning of the immersive reality becoming commonplace.
Machine learning, on the other hand, has already grown by leaps and bounds. Google’s Deepmind, IBM Watson, Microsoft Cortana and the famous Tesla are all products of AI and ML at work.
Technology, Media and Telecommunications Predictions 2018 from Deloitte also confirm that there would be a surge in AI & ML investments.
a) At least one billion smartphone users would be creating AR content in 2018
b) SMEs will invest double the resources for Enterprise Machine Learning development compared to 2017
c) Global smartphone penetration would exceed 90% by 2023.
Put a, b & c together, and it is certain that Augmented Reality and Machine Learning would be used together to solve business challenges and to revolutionize daily lives.
Here are two examples of how AR & ML technologies are being put together to blur the gap between virtual and digital worlds.
Blippar – Making AR + ML Work Better
Blippar is an Augmented Reality company that is harnessing the ability of machine learning to learn and remember imagery. Blippar started off as an Augmented Reality platform for brands and publishers and later expanded its offerings to include visual search.
Image Source: techcrunch.com
Its latest innovation is an automotive identification technology that can identify objects, especially vehicles and recall the make, model name, year of manufacturing of US cars. Blippar has also released a Car recognition API that would enable developers to build on its current abilities.
Pinterest & the Growing Business Of Visual Search
Similar to Blippar, Snapchat and Pinterest have also been expanding massively on their visual search capabilities.
Pinterest released a beta version of Lens, an image recognition tool that enables users to click an image from their immediate surrounding or upload an image from their gallery to search for similar items on Pinterest. Way back in 2015, Pinterest acquired Kosei an ML recommendation engine which connects the dots between user visual search queries and Pinterest’s vast database of images. The pin recommendations as given by the ML engine helps Pinterest increase the time spent by users on its website and mobile app.
AR & ML frameworks on iOS & Android
Apple released CoreML – a machine learning framework that enables developers to add deep learning to apps that run on Camera, QuickType, and even Siri. The unique trait of CoreML is that helps “run machine learning models on the device, so data doesn’t need to leave the device to be analyzed.” read the entire Apple’s official documentation about CoreML.
Apple’s Vision is a framework that facilitates high-performance image analysis to identify faces, objects, images and so on.
Apple ARKit, which was also released with CoreML as part of iOS 11 enables Apple devices to see physical spaces and objects and integrate digital information onto screens with excellent clarity.
Similarly, TensorFlow is the popular open-source machine learning framework that is being used by Google for developing machine learning applications. As for Augmented reality, Google had released ARCore long before the release of iOS 11.
All these frameworks put together will help developers create exciting AR apps with machine learning capabilities. They would throw open a digital world that promises immersive experiences that get better with continuous learning.