Puppy Linux Discussion Forum

Ted Dog · Joined: 13 Sep 2005 Posts: 2051 Location: Heart of Texas

Dependency on an internet connection is hopefully only for the short term. Was able to compile espeak with fatdog. Only need to rename an header file and makefile mod to get it to compile..

Now if I only knew how to construct a 64bit pet. Embarassed

Q5sys · Joined: 11 Dec 2008 Posts: 1046

Ted Dog · Joined: 13 Sep 2005 Posts: 2051 Location: Heart of Texas

thanks for the tip, it works but some how messed up permissions now 'spot' wrapper script is broken (never cared for that needless handicap)

retry 'as spot' to build it.

Odd have 0 upload size. Is that normal? Have I been banned to upload!!

save-as
http://puppylinux.net/fatdog64/espeak-1.47.11d-FD621.pet

ETP · Joined: 19 Oct 2010 Posts: 485 Location: UK

How to change the default "NEWS" site.

Currently when you;
[Tap the HOTKEY and say] “NEWS”
you are taken to the BBC World News page. If you wish the default to be another news page, it is easy to change by simply editing
the following file.(/opt/Palaver-master/Recognition/bin/news)

If for example you wanted it to be Google News:

Original file:

ETP · Joined: 19 Oct 2010 Posts: 485 Location: UK

Teaching Obedient Pup to greet you on startup. EDIT: Do not use with version2

This is just a fun add-on but rather than the usual trick of playing a wav file, uses another of the Google APIs so that
user editable text is spoken. The variety and quality of voices available using this technique is not as good as the
included TTS but is not too mechanical.

1. Using PPM (Puppy Package Manager) install mpg123_1.12.1 from the ubuntu-precise-universe repository.

2. Extract the two files which are attached and reproduced below in your Downloads directory.

3. Ensure that they are both executable then symlink greetings.sh to your Startup folder.

4. Restart the X server and you should hear audio. Edit the two files as you see fit.

greetings.sh

gcmartin · Joined: 14 Oct 2005 Posts: 3617 Location: Earth

I love what this thread's community has done with this with other contributions that was pulled together in this distro. This humanitarian offering, though, is very robust offering to desktop users. I love what you've done. And am looking forward as this product continues its evolutionary path. What is so amazing is that this product provides far-reaching uses to those whom it is designed for, but, it ALSO, provides great services and use to those of us who are not hampered.

This has a great mainstream appeal if more of the community knew about it. It there is a manner of making this appealing to a larger audience, this product could morph into one where more users would mean more heads with ideas for expanding and improving upon the great things it already does.

Thanks to all of you for coming together for the version 1.0 of this effort in Puppyland.

Here to help

ETP · Joined: 19 Oct 2010 Posts: 485 Location: UK

@gcmartin,

Thanks for your kind comments. I am keen to hear what issues testers have found and in particular what percentage success rate they had with the walk-through.
Voice control is not without its problems! Smile

technosaurus · Joined: 18 May 2008 Posts: 4133

It looks like pocketsphinx_continuous is a good light speech recognition tool to complement espeak for text2speech.

One way to handle the output from speech recognition is to use /dev/stdout as the output and pipe it through a while-read-case block like:

gcmartin · Joined: 14 Oct 2005 Posts: 3617 Location: Earth

Thanks @Techno.

ETP · Joined: 19 Oct 2010 Posts: 485 Location: UK

Please note that the included iTunes add-on is currently not working in XBMC as Apple have completely changed their website/service. The author of the add-on “dersphere” is working on a new version and I will post again when it becomes available.
Attempting to play any trailer gives the error message "Remote Share: Could not connect to network server"

Relevant post:
http://forum.xbmc.org/showthread.php?tid=160691&page=15&highlight=dersphere

Git:
https://github.com/dersphere/plugin.video.itunes_trailers

Signing off with:
[Tap the HOTKEY and say] “YOUTUBE Brook Benton Rainy Night in Georgia”

technosaurus · Joined: 18 May 2008 Posts: 4133

I found a decent tutorial/example of using sphinx:
http://hackaday.com/2010/07/11/adding-speach-recognition-to-your-embedded-platform/

ETP · Joined: 19 Oct 2010 Posts: 485 Location: UK

Hi technosaurus,
Thanks for flagging this up. I understand that Ubuntu 13.10 will include speech recognition (PocketSphinx) by default. The final is due out this month.

www.iloveubuntu.net/speech-recognition-engine-pocketsphinx-landed-ubuntu-1310-default

BK has included support for Ubuntu 13.10 (Saucy Salamander) in Woof and we may well see a Pup based on it.

technosaurus · Joined: 18 May 2008 Posts: 4133

built in precise 3.8.3, but you'll still need to populate /usr/share/pocketsphinx/model from the model directory in the source package here:

http://hivelocity.dl.sourceforge.net/project/cmusphinx/pocketsphinx/0.8/pocketsphinx-0.8.tar.gz

Sorry, those parts were too large for a forum post.

ETP · Joined: 19 Oct 2010 Posts: 485 Location: UK

On 10th October I posted about a global problem with this XBMC add-on which is included in this Pup.
It has now been fixed and the add-on will automatically upgrade to version 0.2.1 the next time XBMC is launched.

Signing off with:
[Tap the HOTKEY and say] “YOUTUBE Brook Benton Rainy Night in Georgia”

Atle · Joined: 19 Nov 2008 Posts: 282 Location: Oslo, Norway

is there a simple way to create custom speach commands for someone that does not speak very well?

Such as his way of saying "Click" in order to perform a simple mouseclick?

Great efforts here and lots of kreds to devs. Hope to see this merged with OneSwitch one day Smile

Atle