jessp/holo-assistant

DIY pseudo-holographic virtual assistant

holo-assistant's Issues

Dynamically detect microphone port

Right now, errors happen on pi when the microphone is not in the detected port. Should use sounddevice to find input devices and select first valid option.

Use display for SEPIA avatar?

Hi @jessp ,

this looks like a very cool project 😎 👍 .

Not an issue really but a question. I'm developing an open-source voice assistant as well (SEPIA) and I was wondering if the display might work with the SEPIA avatar (always-on mode) as well 🤔 .
As far as I understand the construction you simply need the animation on a black background right?

Here is a test link to demonstrate one of the SEPIA avatars: https://sepia-framework.github.io/app/index.html?skinId=41&view=aomode (simply click on the little arrow on the bottom of the login box and then "offline test mode" to jump right into the avatar view). Do you think this could work? 🙂

definitions related .wav files are choppy when played on Pi

Files sound fine on Mac but are choppy when played either through Maria (sounddevice) or with aplay. Tried generating the files on sounddevice on the Pi instead but they have the same issue.

Finish documentation website

Precipitation always states it is happening at 0:00

Add wave animation on start up

Add own 3d model

Wanted to know the process you followed to create textures and 3d model.

Don't allow user to start timer if one is already started

If timer is already running, alert user that they must cancel existing timer before starting a new one.

Improve error handling on server

Maria plays talking animation after speech ends

Mostly observed during music-related functions. Most likely caused by sending too many commands in too short order.

Tab with serverAudio/main.py running over SSH becomes unresponsive after running for prolonged periods

Code continues to run, but process is unkillable with "ctrl-z" and equivalents after some minutes.

Upload 3d models for 3d printing

Add python file to force installations of packages on first run

The Spotify API and the places python package require an action that forces the user out of running the code.
This could be made less frustrating by:

• Creating a script that prompts the web-based Spotify API authentication launches
• Creating a script that prompts the geograpy package to install its db

Add clock animation to timer function

Rename skill modules from xmodule.py to xskill.py

This would be a better convention as all things are python modules.

Add definition module

Add module that has Maria defining words

Website down

It appears the website is down

Move keyword listening logic into each skill module

Skills should all encapsulate their own logic rather than relying on main.py to trigger different behaviours.