Home / ubuntu / Questions / 228511
Accepted
Ricardo Altamirano
Asked: 2012-12-13 21:06:13 +0800 CST

Why does the Julius speech recognition engine return a segmentation fault when passing a WAV file?

  • 772

I installed the julius and julius-voxforge packages on Ubuntu 12.04 LTS, and now I'm trying to run the julius command on a WAV file called test.wav. I ran this command:

julius -input raw test.wav

but it simply says "Segmentation fault" without any further information. I've been searching the documentation for a few hours now, and I can't find any relevant information.

segmentation-fault

3 Answers

  1. Best Answer

    I was also able to confirm this behaviour in Ubuntu 12.04

    Your best course of action would be to file a bug, so that you can get proper support from the developers.

    • 1

  2. U cant give input directly on your CLI like this Open terminal on the location where your Sample.jconf is present(Where your Acoustic Model and Grammer is present) Then use this command

    julius -input rawfile -C Sample.jconf

    After that input your wav file with its location Remember the input file should be in- 1) .wav format 2) bit resolution - 16 bit 3) sampling rate - 16000 hz 4) audio channel - mono

    • 0

  3. I think that you forgot to call configuration file. Like: -C config

    • -2