ideas.md

for tile lights

• conway game of life
  • add transitions between states
• arkanoid
• flip through images
• breaking news/weather
  • font-to-display parser

for demo:

• fix misty's sounds/moods

• fix train face - test with lower volume?

• calibrate misty

• to demo:

  • audio:
    • say
    • play
  • movement
    • search?
    • nod?
    • shake_head?
  • touch
    • respond to touch
  • compose
    • smooth_jazz
    • party_mode
    • lets_get_silly
  • train_face
  • from utils.misty.routines.identify import run

• to discuss:

  • i wrote the API i'm using
  • i got to use a field trial version
  • they out to interview me
  • dialogflow - for conversations with xaxis traders
  • CES: recognizing faces and saying hello
  • CES: responding to questions about schedule
  • CES: mistyrobotics will be there and are willing to help out and they interviewed me at my place

• concerns with interaction (from kelley email):

  • volume (use a bluetooth connected speaker)
  • keyphrase (might be too loud - instead respond to touch?)
  • recording audio (workaround discovered - lower the volume to around 50)

bugs

• add ignore to reset_to_orig
• can't set both arm positions simultaneously, for some reason

for WIT

• simple mad libs, perhaps with different voice
• face training/recognition:
  • improve on challenge
  • have misty search for people
  • say hi to them as they walk by
• internet test
  • use wifi/cellular from usb-tethered phone
• generic/easy import file
  • from utils.misty import *

to bring:

• misty
• macbook
• chargers
  • phone
  • macbook
  • misty
• power strip or 2
• extension cord
• hdmi cable
• raspberry pi?

ideas

• simple dialogue flow to determine yes or no
• grab seats from database and say hello in all languages
• create a keyboard reader
• youtube to mp3: https://mp3converter.to
• text doc of all things to say at WIT

config for tts audio for misty:

        {
          "audioConfig": {
            "audioEncoding": "LINEAR16",
            "pitch": 2,
            "speakingRate": 1
          },
          "input": {
            "text": "hi, i'm co-pi-lette"
          },
          "voice": {
            "languageCode": "en-US",
            "name": "en-US-Wavenet-E"
          }
        }