Voice assistants or personal voice assistants are programs using natural language processing (NLP) and speech synthesis to perform certain tasks on the user command. Currently, they have been a major part of our smartphones, computers for the past few years, If you are an iPhone user certainly you are using Siri, or if you are an android user you know your Google Assistant.
Voice assistants had been released by the major players in the past few years as Apple, Microsoft, Amazon, Google, and Facebook. Now we have open-source projects giving the choice to individuals to create their own personal assistants and implement it in their smart homes.
What is the difference between voice assistant and intelligent assistant?
Intelligent assistant is a program built to assist the user to perform tasks, It often uses natural language processing and artificial intelligence. However, voice assistant is an intelligent assistant itself but technology oriented as it uses voice recognition, speech synthesis alongside with the natural language processing(NLP).
How does Voice assistant work?
Voice assistant respond to a certain keyword as "OK Google”, "Hey Siri", "Alexa" and "Hey Cortana", activate the assistant to await further instructions or task
Voice assistants are a production of integration of AI into our daily routine, They are intelligent enough to recognize their user voices and some are built to perform complex tasks based on machine learning algorithms.
What voice assistant can do?
- Taking notes
- Creates a calendar
- Booking appointments
- Calling a phone
- Searching the internet
- Creating reminders
- Make Reservations
- Search a location
- Reading notes, books and articles
- .... and more.
Open-source technologies give flexible options to creators, startups and developers to build their own consumer products. That allows the developers to experiment, expand the usability to another field and build sector oriented products.
Mycroft is an open-source voice assistant built for Linux machines, can be installed on Linux desktops and Raspberry Pi. Mycroft is a privacy oriented which doesn’t collect and monetize your data.
Mycroft AI has released Mycroft Mark 1 which is Mycroft enabled device aimed for developers as an advanced prototype, Mycroft Mark II is due to be released this month (December 2018), which is a consumer-ready speaker for end-users.
Mycroft has provided developers with documentation to build their products using Mycroft on Linux-based device so the developers are not limited to Raspberry Pi but any Linux ready machine and single board computer.
Currently, Mycroft has no support for Windows and Mac OSX, I subscribed to their mailing list to be notified about Mac OSX and Windows support for more than a year not. Hopefully, other operating systems will get supported soon.
- Platforms: Linux, Raspberry Pi
- Developer friendly : yes
- Development status : active, community based
Kalliope is an open source modular based voice-controlled personal assistant designed specifically for home automation. It was built to run on Linux, Raspberry Pi, and Android phones.
Kalliope is built for developers with powerful rich developer-friendly documentation, API and tools.
Kalliope marketplace ecosystem.
As Kalliope is built on a modular architecture, It has a marketplace for signals which trigger the response, Neurons which are plugins to perform specific actions. Neurons marketplace has core neurons and community neurons by community developers.
More developer tools:
- Kalliope REST API
- Kalliope voice assistant interactive shell
- Kalliope Web UI
- Kalliope android application
Platforms: Linux/ Raspberry Pi and Android
Stephanie is an open-source platform built specifically for voice-controlled applications as well as to automate daily tasks imitating much of a virtual assistant's work.
Stephanie voice assistant has been released under MIT license, with rich documentation intended for end-user and developers to install, configure and use it. The developers can easily extend it by creating their modules.
Stephanie code has been stared about 700 times and forked about 100 times.
Platform: Linux, Windows and Mac OSX
Developer friend: yes
Development status: stable
Open Assistant is an open source personal assistant. It responds to voice commands in a conversational rich dialog as in the demonstration video. It has been in the prototyping phase for so long, yet it manages to be in the watch for developers and as it has been forked many times.
Open Assistant is easy to install, However, it lacks the developer documentation. Considering there are only one core developers, the project is still in active development.
Platforms: Windows, Linux, and MacOSX
Dragonfire is a virtual assistant built specifically for Ubuntu-Linux distributions. It provides an Android client that can be installed on Google Play Store.
Dragonfire uses Mozilla DeepSpeech to understand your voice commands and Festival Speech Synthesis System to handle text-to-speech tasks.
- Platform: Linux (KDE neon, elementary OS, Ubuntu )
- Language support: English
Jasper is an open source platform to create voice assistant and extend the application with voice assistant features. However, It's Raspberry Pi oriented so far.
Jasper has been released under MIT license, It provides rich documentation, and developer friendly API to build applications.
Platform: Raspberry Pi and Linux
Jarvis is an open-source assistant, It can be installed and run through the terminal of Linux, Mac OSX and Windows. It has a modular structure. It has voice assistant features. Though the project is still in active development. It's mainly for developers and advanced users in the current state.
Platforms :Linux, Mac OSX and Windows
Linto is an open-source voice assistant, built for business. It focuses on automation the meeting tasks. Linto has an open-source device aiming to provide a customizable platform for developers, smart speakers manufactures, and replace desk phone hardware.
Linto team has released all of its tools as an open source at GitHub. Yet there is no central documentation published yet.
Jovo is a yet another open-source platform to build voice-enabled applications on top of Amazon Alexa and Google Assistant. With one code base, The developer can create their application and deploy it with ease.
Jovo has provided the developers with everything they need as open source tools. There is an enterprise paid plan includes custom features and plugins to facilitate the production of the application with a set of development tools.
Platform: Google Assistant and Amazon Alexa
Hey, Athena Is a free open source modular voice assistant framework. It was built for developers. However, Hey Athena website is down, so as documentation and the code has not been updated for about a year, Though the project code is available for developers.
Project description: a framework
Note: This article is an Evergreen article. It will receive continuous updates of the listed applications. You can save it and return to it as an updated resource.