Relevant links:
Mobile Application |
ESP8266 Smart Light Demo
- awarded Gold in Foundation/Equivalent Category
- does not use servos anymore
- Eyeball is now done through a screen and PyQt (interface)
- app camera stream and eye movement can work simultaneously
- utilizes Gemini (previously ChatGPT), both Gemini Pro and Gemini Pro Vision
- Text Generation is faster
- can now translate what it sees from the camera using Gemini Pro Vision
- serve certain parts on the cloud (AWS or Azure) for remote controls
- passively listen for wakewords
- improve mic pickups
- awarded Silver in Category A
developed an app (unfinished)
pairs with a smart light as demo to test home automation
fully embedded on a Raspberry Pi 4B
is 3D printed (PETG+)
camera is now inside the eyeball itself
camera acceleration (using quadratics)
- Allow users to register as "Friendly Members", whereby any motion detection on the "unregistereds" will send off an intruder alert to the app
- Add status box for paired smart products (Time elapsed, Power Consumption)
- Use only one powersupply
- Resolve the servo power issue
- Get better speaker
User-tracking robotic eyes hooked up with ChatGPT (as brainpower) to serve as generic personal assistant robot
- Won 3rd place in UITM's Arduino Showcase 2023 (cash-prize of RM100)
- Microphone does not pick up wakewords in a loud environment
- Goes crazy when there is more than one person detected on tracker
- Bad build quality
- Not a fully embedded machine
- Does not have an actual monetizable use case (aside from being able to track and talk)