Uses the Brilliant Labs Frame camera to take a picture that includes some text, uses Google ML Kit text recognition then translation to translate the text and display it on Frame. All processing is done with local on-device models on the phone and takes about 3 seconds from triple-tap to display.
This demo currently only performs Japanese to English translation, although other script is possible. (Some screenshots show images from when Chinese/English was specified.)
framecast1.mp4
