Somewhat fancy voice command recognition software
Vous ne pouvez pas sélectionner plus de 25 sujets Les noms de sujets doivent commencer par une lettre ou un nombre, peuvent contenir des tirets ('-') et peuvent comporter jusqu'à 35 caractères.
Clara Hobbs 6347c5fdbb Replaced last YAML bit with JSON il y a 8 ans
assets Added Gtk Tray UI (thanks padfoot) il y a 10 ans
data Added Gtk Tray UI (thanks padfoot) il y a 10 ans
.gitignore Ported to new pocketsphinx and pygi il y a 8 ans
README.md Replaced last YAML bit with JSON il y a 8 ans
blather.py Replaced last YAML bit with JSON il y a 8 ans
commands.tmp added gstreamer0.10-base-plugins to the requirements in the README il y a 10 ans
gpl-3.0.txt licensing il y a 11 ans
gtktrayui.py Beginning work moving from YAML to JSON il y a 8 ans
gtkui.py Beginning work moving from YAML to JSON il y a 8 ans
language_updater.sh The updater uses the correct commands file il y a 10 ans
options.json.tmp Beginning work moving from YAML to JSON il y a 8 ans
recognizer.py Beginning work moving from YAML to JSON il y a 8 ans

README.md

Kaylee

Kaylee is a somewhat fancy speech recognizer that will run commands and perform other functions when a user speaks loosely preset sentences. It is based on Blather by Jezra, but adds a lot of features that go beyond the original purpose of Blather.

Requirements

  1. pocketsphinx
  2. gstreamer-1.0 (and what ever plugin has pocketsphinx support)
  3. gstreamer-1.0 base plugins (required for ALSA)
  4. python-gobject (required for GStreamer and the GTK-based UI)

Note: it may also be required to install pocketsphinx-hmm-en-hub4wsj

Usage

  1. Move commands.tmp to ~/.config/blather/commands.conf and fill the file with sentences and command to run
  2. Run blather.py, this will generate ~/.config/blather/sentences.corpus based on sentences in the ‘commands’ file
  3. Quit Kaylee (there is a good chance it will just segfault)
  4. Go to http://www.speech.cs.cmu.edu/tools/lmtool.html and upload the sentences.corpus file
  5. Download the resulting XXXX.lm file to the ~/.config/blather/language directory and rename to file to ‘lm’
  6. Download the resulting XXXX.dic file to the ~/.config/blather/language directory and rename to file to ‘dic’
  7. Run blather.py
    • For GTK UI, run blather.py -i g
    • To start a UI in ‘continuous’ listen mode, use the -c flag
    • To use a microphone other than the system default, use the -m flag
  8. Start talking

Note: to start Kaylee without needing to enter command line options all the time, copy options.json.tmp to ~/.config/blather/options.json and edit accordingly.

Bonus

Once the sentences.corpus file has been created, run the language_updater.sh script to automate the process of creating and downloading language files.

Kaylee now updates the language automatically. You should never need to run language_updater.sh manually.

Examples

  • To run Kaylee with the GTK UI and start in continuous listen mode: ./blather.py -i g -c

  • To run Kaylee with no UI and using a USB microphone recognized and device 2: ./blather.py -m 2

  • To have Kaylee pass the matched sentence to the executed command: ./blather.py -p

explanation: if the commands.conf contains: good morning world: example_command.sh then 3 arguments, ‘good’, ‘morning’, and ‘world’ would get passed to example_command.sh as example_command.sh good morning world

  • To run a command when a valid sentence has been detected: ./blather.py --valid-sentence-command=/path/to/command note: this can be set in the options.yml file
  • To run a command when a invalid sentence has been detected: ./blather.py --invalid-sentence-command=/path/to/command note: this can be set in the options.yml file

Finding the Device Number of a USB microphone

There are a few ways to find the device number of a USB microphone.

  • cat /proc/asound/cards
  • arecord -l