When recognizing specific spoken words / faces / text, play a theme song or video, with a little randomness.
Themes needs a TIM, a device with speakers, a screen and a webcam and can run standalone. Themes can be triggered via things that TIM recognizes or via the API.
A bit of video, audio, image or text
- Faces
- Words
- Who is near TIM
- Things TIM said earlier.
- Speakers: play an audio file
- Screens: display an image or video
Add an event for TIM to recognize: Message:{creator, name, type, recognize, theme} REST /api/v1/themes
An external process or detached TIM, a TIMd, could also want to trigger a theme, this could be done by calling: Message: {event: recognize} REST/api/v1/notify
TODO: add raw audio/video/text stream input TODO: Authentication, connect to a specific TIM.io -> return key