Created clone of Siri - Sirius, available under the BSD license

Group Clarity Lab , located at the University of Michigan, offers a solution to the open source personal computer assistant. Their program recognizes the voice and looking for answers to questions, just like Siri make from Apple, Google Now from Google, and similar solutions from Microsoft and Amazon. Project called Sirius knows even more than his counterparts - for example, it can provide a picture and ask for her question.

The project is sponsored by Google, DARPA, the US military and the US National Science Foundation. It spreads by лицензии BSD , which means full open and free for everyone. This means that any start-ups can implement such functionality in their designs and devices.



The system works i>

But for this they have to understand the rather complicated system, which consists of the project, laid out on GitHub . Operation of the system requires additional individual programs - Sphinx , Kaldi, Protobuf, Speeded Up Robust Features (SURF , is based on OpenCV). The conference ASPLOS 2015 as developers will present their scientific work dedicated to the project, and tutorial on installing and working with him. < br />
Among enthusiasts have been trying to create an analogue of Siri - it is, for example, JuliusJS , library for speech recognition. But while significant results from these projects are not visible.

Source: geektimes.ru/post/247150/