Facebook has released Caffe2, which is an open-source platform that will enable to bring AI directly to our smartphones and tablets.
Having artificial intelligence on our mobile devices is becoming commonplace these days. Many of our modern smartphones are equipped with AI assistants. Apple has Siri, Google has Google Assistant, and Samsung’s new lineup of S8 smartphones will see the debut of their new virtual assistant, Bixby. Amazon Alexa is also coming to the iPhone. However, Facebook has now announced the launch of their new open-source platform, Caffe2, which will provide coders with the machine-learning tools necessary to develop AI software for mobile devices.
This isn’t really anything new. Google’s TensorFlow for instance, has been around for a while now, and it basically offers the same deep learning tools as Facebook’s Caffe2. However, where Mark Zuckerberg’s company’s open-source platform differs is that it offers, as Facebook stated in a blog post, “a lightweight and modular deep learning framework emphasizing portability while maintaining scalability and performance”.
What TensorFlow does is use the cloud to power the phone’s AI software, which just slows things down. Caffe2 eliminates this barrier completely, by harnessing the power of the mobile device’s own hardware in order to speed up the AI’s software. This means that users will be able to have image recognition, natural language processing, and computer vision directly on their smartphones.
“We’re committed to providing the community with high-performance machine learning tools so that everyone can create intelligent apps and services,” wrote Facebook.
In order to ensure that Caffe2 is able to run efficiently both on the cloud and directly on mobile devices, Facebook worked closely with NVIDIA, Qualcomm, Intel, Amazon, and Microsoft. Facebook hopes that by collaborating with these world leaders in the technology space, people will experiment with Caffe2 and therefore develop the next generation of AI software.