In fact, Apple has a machine-learning application that is full of two pages. Some of them have already been put into use. Others are still under discussion. Behind these applications is that Apple has quietly acquired 15 artificial intelligence companies over the past 6 years, including innovative technologies such as speech recognition, natural language processing, image recognition, face recognition, motion capture, and machine learning. the company.
An interesting phenomenon is that after each acquisition of an AI company, Apple will not publicly announce the acquisition purpose and development plan, and will immediately close the products and services of the acquired company.
Speech, semantic recognition
Siri Voice Assistant
Siri is undoubtedly Apple's most important voice product. Buying it at a price of US$200 million in 2010 was undoubtedly a very good deal. Afterwards, Apple’s acquisitions in the direction of voice recognition and NLP were all based on Siri.
Novauris speech recognition
In April 2014, Apple acquired Novauris, an automated speech recognition technology company. Novauris was founded by Nuance, the founder of DragonSystems, and Nuance was the basic voice technology provider for Siri. Their core product is the server-based NovaSystem, a scalable speech recognition system that features simultaneous processing of multiple voice access requests. Although Apple declined to disclose how to use the Novauris team, from the perspective of the relationship between Siri and Nuance, Apple's acquisition is intended to get rid of Nuance's dependence and use the Novauris team to develop its own voice technology.
Vocal IQ human-machine dialogue
All along, Siri can only identify the most basic instructions and can't make high-matching answers, which makes it an entertainment product for entertainment. In October 2015, Apple resolved this issue by acquiring VocalIQ. VocalIQ can use deep learning to understand the language environment and make human-computer conversations more natural. After embedding the artificial intelligence technology of VocalIQ in Siri, the developer uses this platform to store and learn user's communication information, to accurately identify the user's instructions and provide more intelligent dialogue.
Machine learning
Apple once mentioned that they use AI technology to test some micro-functions: such as identifying strange calls; detecting the user's movement status; listing the most likely to use applications after the phone is unlocked; automatically displaying nearby marked hotels; scheduling; interest news recommendation and many more. These seemingly simple functions were all put into the machine learning system. For this reason, Apple used its machine learning company, which was interested in recommendation and habit prediction, to improve its data mining capabilities.