el-tocino/localcroft

Bits for locally-served Mycroft instances

localcroft's Issues

Create config-level entry to enable filtering items.

Within the mycroft conf you should be able to add local config entry to enable/disable denoise/highpass/lowpass/normalize/other? filters. Right now they're all just on by default in local deepspeech server config. They could be used on other engines as well, which I have not tested with those adjustments, if one was so motivated to try*. Performance considerations might also change what people want to have filtered. The high/low pass frequencies would also be nice things to allow config settings for.

• this is left as an exercise for the reader.

Find a way to avoid writing output temp file

localcroft/mycroft-core/mycroft/stt/__init__.py

Line 307 in 02c8d8a

ds_audio.export("/opt/mycroft/stt/ds_audio.wav",format="wav")

localcroft/mycroft-core/mycroft/stt/__init__.py

Line 308 in 02c8d8a

stt_audio = open("/opt/mycroft/stt/ds_audio.wav", "rb")

There's got to be a way to skip writing a temp file just to read it back in.

Question about wav content

Hi el-tocino,

I'm struggling a bit to find a german dataset to speed up the process of finding fake words.

There are some sets, but almost exclusively spoken sentences (half-sentences). Some are short, but i'm not certain that this even qualifies to be training material. Is precise-train-incremental restricted to spoken words?

Question about local wikipedia with ZIM and Kiwix-server

Hi El Tocino

Thank you very much for you guide on all things local mycroft.
:)

Re. Wikipedia, I followed the easy method with Invader ZIM.
All went well and I can access my local wikipedia via web browser.

I am having issues getting mycroft to access this wiki though.
I edited the file wikipedia.py and changed the API_URL with the local IP (http://192.168.1.43:7998/) of the kiwix-serve:
When querying mycroft wikipedia skill, I get the error WikipediaSkill | Error: Expecting value: line 1 column 1 (char 0)

Would you mind please sharing your config of the API_URL in wikipedia.py?

Many thanks.

[precise] Wakeword improve activation speed

While I started writing a rather chunky issue over on the mycroft-precise repo, I re-read your documentation and though I'd better ask here/you directly.

What I'm currently struggling with is activation speed. I've been very carefully in regards to my training data and ensured every clip starts immediately with the wake-word followed with 1 second of silence (silence meaning quiet room -> me not speaking).

Using my dataset combined with the following for not-wake-words:

• http://pdsounds.tuxfamily.org/
• https://github.com/MycroftAI/Precise-Community-Data/tree/master/not-wake-words/noises

reaches a val_acc 1 in about 120 epochs (super quick).
While it activates quite consistantly, it does so rather slow as it requires for the trailing 1 second to pass as well.

If I now duplicate the data-set and strip 500 ms from the end of every single wake-word clip, I'm suddenly unable to reach a val_acc higher than 0.5
Stripping 800-1000ms has me sitting on val_acc 0.
Training for more epochs (I tried up to 6000) did not help.

Is this to be expected? Is there a way to work around this?
Any help would be much appreciated and thanks for your current write-up. It already helped a lot :)

graphics

hello,

I'm new to using precise, I read your file https://github.com/el-tocino/localcroft/blob/master/precise/Precise.md, and feel very useful. but how can you get the graphics from training ?

thanks