Skip to content

Latest commit

 

History

History
78 lines (61 loc) · 2.9 KB

ubuntu_install.md

File metadata and controls

78 lines (61 loc) · 2.9 KB

Installation on Ubuntu and related Debian based distributions

Tested on:

  • Ubuntu 16.04 (2 August 2017)
  • Ubuntu 17.04 (2 August 2017)
  • Ubuntu 18.04 (10 July 2018)
  • Linux Mint 18.3 (10 July 2018)

Additional steps required for Ubuntu 18.04

  • Java is not installed by default and there are some compatibility issues with the latest version of Java (Java 10 as of writing) installed by default-jdk therefore, you need to manually install an older version of Java (tested with OpenJDK Java 8).

    • You may install OpenJDK's Java 8 by running:
    • $ sudo apt install openjdk-8-jre openjdk-8-jdk

  • There are some issues with ca-certificateson Ubuntu 18.03 (as of writing) and you will encounter some errors while building susi_server if not rectified

    • You will need to run the following commands:
    sudo echo '\xfe\xed\xfe\xed\x00\x00\x00\x02\x00\x00\x00\x00\xe2\x68\x6e\x45\xfb\x43\xdf\xa4\xd9\x92\xdd\x41\xce\xb6\xb2\x1c\x63\x30\xd7\x92' > /etc/ssl/certs/java/cacerts
sudo /var/lib/dpkg/info/ca-certificates-java.postinst configure

Steps

  • Clone the Github Repository and open folder
$ git clone https://github.com/fossasia/susi_linux.git
$ cd susi_linux
  • Run the install script
$ ./install.sh
  • Verify that your Audio setup is done properly. For this, first we need to check for recording devices. Run command
$ rec a.wav

Verify that it gives an output like below.

Ubuntu Rec Command

  • After this, play your recorded audio by running play a.wav. It should give an output like below and your audio must be audible to you.

Ubuntu Play Command

If you hear your voice properly and output is similar to what shown in screenshots, setup is done correctly. If you face an error, try running pulseaudio -D and re-running the commands. If still there is error, see if devices are selected correctly in Ubuntu sound settings.

  • Run the configuration generator script and optimize the setup according to your needs.
$ python3 config_generator.py

-Note: Enclose every input(y/n, email, password) queried after running the above command with ''(single quotes).

  • One config.json is generated, you may run SUSI User Interface by executing the following command
$ python3 app.py
  • Alternatively ,you can run SUSI without User Interface by executing the following command
$ python3 -m main

In both case SUSI will start in always listening Hotword Detection Mode. To ask SUSI a question, say "Susi". If detection of hotword is successful, you will hear a small bell sound. Ask your query after the bell sound. Your query will be processed by SUSI and you will hear a voice reply.

Faced any errors?

If you still face any errors in the setup, please provide a screenshot or logs of errors being encountered. This would help rectify the issue sooner.