Have you heard of Mycroft Mark II?
Smart speakers and virtual assistants have become a major consumer-electronics category since 2018. These devices not only allow you to control your surroundings just by your voice command. But they have the potential to do much greater things. These speakers are capable of hearing each and every conversation made in its proximity. The company collects your data unless you choose to delete it or turn off the microphone of the speaker. Companies like Amazon, Google, and Apple are some leading smart speaker manufacturers. All of them have been reported to hear those conversations without their customer’s permission.
Well, this doesn’t mean that you shouldn’t buy one. If you are looking for a smart speaker personal assistant that maintains data independence and privacy then Mycroft Mark II can be taken into consideration.
What is Mycroft Mark II?
Just like any other smart speaker, Mycroft Mark II works the same way. But it also authorizes its users to safeguard their personal data and privacy.
This wireless open-source virtual assistant can perform tasks like :
- Playing and streaming music.
- Setting up events and accessing pre-existing events from the calendar.
- Conduct an internet speed test.
- Connect with other devices over WiFi.
- Answering questions and a lot more.
These tasks are very basic when it comes to virtual assistants. They can also be done by other virtual assistants as well. But the special features that make Mycroft Mark II different from other virtual assistants are :
- It is a privacy-oriented smart speaker that focuses on user privacy the most by deleting the recordings as soon as the results are given.
- Mycroft uses next-generation machine learning to collect the data from the user(which can be disabled). Just like for speech-to-text transcription, language understanding, speech synthesis and to improve spotting of the wake word.
- One of the coolest features of all is the ability to customize the wake word. The wake word is a special word which when spoken, activates the device. Mycroft Mark II allows its users to modify the voice and even the user experience.
- Mycroft has got a completely open software, hardware, and data platform. This simply means that the users can customize their experience with this device according to their wills.
- The design of Mycroft Mark II is very elegant. Mycroft has made the Mark II as a result of “non-stop iterations” by analyzing all the feedbacks and community contributions of its predecessor- Mark I.
Also Read : DIY Mycroft on Raspberry Pi 4
Main Aim Of Mycroft Mark II
The main aim of Mycroft to build the Mark II was to provide its user with an option not to store their voice data into its servers and to be more transparent to their users, unlike the other voice assistant manufacturers. The company even mentions that the user data would not be sold to any advertisement brands or to any other sources making it the only voice assistant out there that truly respects user privacy.
This was just a basic overview of the Mycroft Mark II. Now let us check the specifications of the Mark II.
Specifications of Mycroft Mark II
The dimensions of Mycroft Mark II are
- 7.72 x 4.2 inches
- 196mm x 105 mm (height x diameter)
making it a good compact smart speaker that can be kept at any place of the house. It is equipped with an impressive 10-watt speaker of two inches full-range drivers that produces loud and crisp stereophonic sound.
The Mark II has also got an array of six individual microphones in all directions that are capable of identifying the user voice direction using noise interference cancellation, acoustic echo cancellation, and beamforming technologies. No matter how chattering or noisy the environment be, Mycroft Mark II will be able to isolate the users’ voice.
Mycroft Mark II has also got a 4-inch IPS LCD touchscreen display in portrait form. The LCD touchscreen panel can display information, some animated graphics of clocks, calendars, and other stuff, and also a friendly robotic interactive face and a simple user interface. The screen can play content from Google’s video services as well. Since the voice assistant is built on an open-source AI, it can even be further modified or customized by the user but for this, some good coding skills are required.
Connectivity Of Mycroft Mark II
This smart speaker is also WiFi equipped supporting 2.4 GHz and 5G connectivity for connecting with other devices over the internet (IoT). Also, it has a Bluetooth connection available along with a microSD card slot for easy software updates, and a USB Type-A port for file and I/O transfer. Along with English, this voice assistant will also be supporting Spanish, Italian, French, German, and Portuguese as well.
Last but not the least, all the skills of Mycroft Mark II are powered by a 16 nm Xilinx Zynq UltraScale+ MPSoC. It comes with two 64 bit Arm Cortex-A53 processors. It also comprises a 3.5mm audio out and an 18- watt power supply with international adapters.
- Xilinx quad-core processor
- Far-field 6-microphone array
- Hardware AEC, beamforming and noise reduction
- Stereo sound with dual 2″ drivers (10W)
- 4″ IPS LCD touchscreen
- USB Type A
- MicroSD card slot
- Bluetooth in
- 3.5mm audio out
- 18W power supply with international adapters
Is Amazon’s Development Kit better?
If we have a look at Amazon, it uses a virtual assistant AI technology called Alexa. The Amazon Alexa Premium Far-Field Voice Development Kit allows developers to integrate Alexa on their products. Just like Mycroft, the products built on the Amazon Development kit are provided with a Far-Field Voice Recognition system.
It comes with an array of seven microphones arranged in a circular layout. Also, an array of eight microphones in a rectangular configuration which is also seen in Echo and Echo Show.
Amazon’s dev kit also uses audio processing algorithms and Wake Word Engine (WWE) that run on Intel’s dual DSP with inference engine.
This can be compared to the Mycroft Mark II development kit that provides us with a precise wake word spotter and DeepSpeech STT by Mozilla. Mycroft’s text to speech synthesizer (called Mimic) uses machine learning to synthesis voice on its devices.
It uses Machine learning in every course of interaction and it is all on an open-source platform. Also, the data collected from the user is deleted as soon as the interaction is completed unless the user allows it. The customizable wake word is also an additional feature that makes this open-source platform unique. Therefore, both these development kits have their own advantages. Mycroft Mark II has 2 64 bit Arm core + DSP +fpga for far-field audio signal processing.
Far Field Microphone Technology
Amazon has partnership with many vendors who provide microphone array and far field microphone technology to be integrated with Amazon Alexa.
Mycroft Virtual Assistant
The Mycroft Mark II smart speaker uses a Mycroft intelligent assistant which is an open-source platform. Mycroft is the world’s first open-source voice assistant. It is largely based on the Python programming language. Being an open-source platform allows users to play around or contribute to the project. The fact that they offer a very wide range of development on the basis of consent, distinguishes Mycroft from other popular assistants.
Mycroft has a partnership with Mozilla’s Common Voice Project (called Myzilla) to grip their DeepSpeech speech-to-text software especially build for their Mark II. Development on an open-source platform like Mycroft AI takes time and needs to be built carefully but the final output results are very convenient.
Since Mycroft is open-source software, it can be run on many hardware, GNU/Linux computers, or Android phones. Mycroft has an image for Raspberry Pi 3 or 4 called Picroft.
Being an open-source virtual assistant Mycroft is very good but it has got some rough edges that are being taken care of. Major companies that make virtual assistants like Amazon, Google, and Apple are more interactive and provide you with amazing and creative answers but Mycroft’s answers would not be as appealing as theirs to every consumer.
The next topic we’re gonna talk about is How different the Mycroft Mark II is from the other standard voice assistant smart speakers.
Is Mycroft Mark II better than Mark I other smart speakers?
The first version of Mycroft smart speaker virtual assistant; Mark I was more focused on developers and programmers whereas its successor, the Mark II is more like a consumer-ready product but at the same time, it runs on an open-source software allowing the developers to customize and change its configurations. Therefore, Mark II is more user-friendly than Mark I. Mycroft Mark II also offers the better sound quality, improved voice detection, and is more aesthetic than Mark I.
When we compare Mycroft Mark II with any other virtual assistant smart speaker like Google Home, Amazon Echo, or Apple HomePod, we notice that Mycroft is not consumer-ready yet though it is developing at a tremendous rate. The reason why Google, Alexa, and Siri are so good at understanding the user is that they store each and every single word spoken within their reach without the user knowing about it and used to improve their speech recognition.
These companies also sell this data to advertisement brands and other sources to earn a significant amount of revenue out of the data they collect from their users. These things do not happen with Mycroft Mark II since it does not save your data on a third-party server. It does send your data to a cloud server to process requests. But the only difference here is that the requests are deleted right away.
Your conservations with Mycroft are very private and you won’t be seeing any ads on your socials regarding your talks. The main aim of Mycroft with Mark II was to offer a much more customizable and programmable smart speaker than the Echos and HomePods. Therefore, Mycroft Mark II offers much more than just privacy.
Wrapping it Up: Should you consider buying Mycroft Mark II?
The Mycroft Mark II has got some appreciable top-grade features like a
- bright touchscreen display,
- customizable wake word
- array microphones,
- and the most importantly, privacy which most of the smart speakers lack.
Mycroft Mark I was specifically made for developers. Mycroft has made some significant changes in the Mark II that target regular people as well.
Considering the price of Mark II $129 and the features and the absolute privacy being offered, Mycroft is a pretty decent deal. If you are a developer and crave to innovate something new along with a virtual assistant, then Mark II might be the product you are looking for since no other smart speaker runs on an open-source virtual assistant. But if you are a normal consumer who is looking for just a very simple smart speaker and privacy does not greatly matter to you, then probably this product won’t amaze you much.