Bless You Polar Bear

Front of 'Bless You Polar Bear'
Back of 'Bless You Polar Bear'
LCD Display
Moblie Microphone

1. Description

In Western culture, saying "Bless you" after a sneeze is a warm social gesture. However, for those from different cultural backgrounds, this habit can feel unfamiliar. To bridge this gap, we created "Bless You Polar Bear"—an interactive device with a square LCD and speakers. When it detects a sneeze, it instantly offers a friendly "Bless you!" and context-aware health advice based on real-time weather. Also, It serves as a cultural bridge and a mental-care companion for international students and anyone living alone.

2. Inspiration

Our team consists of Korean students in the U.S. who often felt awkward after sneezing because we weren't used to the "Bless you" culture. After hearing that some locals feel lonely when no one responds to their sneeze, we realized this small phrase carries significant social weight. We wanted to create a device that helps people adapt to new environments while providing comfort to those who feel lonely.

3. What it does

Sneeze Detection: Uses a custom-trained machine learning model to detect sneezes in real-time.
Visual and Audio Response: Displays a "Bless You" animation on the LCD and plays a warm voice message.
Smart Health Advice: Fetches local weather via Wttr.in and uses Gemini API with ElevenLabs TTS to provide context-aware tips (e.g., pollen warnings) right after the greeting.

4. How we built it

We used a Raspberry Pi connected to a square LCD, speaker, and moblie phone(microphone), all housed inside a polar bear. The software stack includes a custom-trained ML model for sound recognition, Gemini API for generating natural advice, and ElevenLabs for high-quality audio.

5. Challenges we ran into

Initially, the model mistook finger-snap, clapping or loud talking for sneezes. Also, it failed in noisy environments. To solve this, we built a custom dataset mixing sneezes with ambient noise and included "negative data" like coughs, throat clearing, and chair-scraping sounds to improve accuracy. We also resolved Raspberry Pi power issues by switching to a proper 3A cable.

6. Accomplishments that we're proud of

We successfully developed a robust, real-time sneeze-recognition model that can distinguish a sneeze from various daily background noises.

7. What we learned

In a time-limited hackathon, quick communication and agility are more vital than a rigid plan. We also learned to treat every trial and error as a necessary step toward a better solution.

8. What's next for Bless You Polar Bear

Form Factor Evolution: We plan to make the device more huggable and portable, allowing users to carry their companion anywhere.
IoT Integration: Using on-device detection for privacy, we aim to trigger smart home devices like air purifiers or fans upon sneezing for better health management.
Public Health: Applying this tech to public facilities to monitor sneeze frequency and automate ventilation for infection control.