TODO

• electronics
  • add PSRAM to board
  • add I2S breakout to board
  • assess power usage and increse efficiency
• firmware
  • Displays search engine results (demo usage)
  • Mic + Display simultaneously (I2S mic on I2S channel 1, display on I2S channel 0)
  • Contextual search
  • Live language translation
  • Live captioning (native language)
  • Receive images + render using LVGL
  • Update to transmit data over Bluetooth
  • Fix memory leak on data received (probably with cJSON)
  • App support, refactor codebase with apps in mind (reference InfiniTime repo)
  • Support OTA updates (may need to reduce flash usage)
• mechanical
  • test out current cad model - wear around with lenses
  • make battery cavities + battery wire hole in cad model
  • make hole for external microphone in CAD model
  • slim down top bar/top rims of front frame
  • display version - make arms bigger to accomodate 14.5mm wide driver board
• get microsoft for startups or aws activate
  • use aws activate or move from AWS to azure for startups
• add saving of transcripts to mongo database with user id
• software
  • fix WIS + OSSG so that transcripts show up live again and we don't return intermediate transcripts over Wifi many times per second - larger packets or some buffer changes
  • contextual search
    • change to take last n words instead of just current transcript in api ping
    • add word definition functionality for low-frequency words
    • add creating of users with web frontend
    • add location/map functionality with NER and maps API
• changing WIS and OSSG to use only Bluetooth for all communications