Skip to content
VibeOut

VibeOut

Unlock Your Best Self: Emotion-Powered Fitness and Support with GenAI

Created on 24th September 2024

VibeOut

VibeOut

Unlock Your Best Self: Emotion-Powered Fitness and Support with GenAI

Describe your project

Our project is an Emotion-Based Workout and Support Assistant that uses AI to enhance users' mental and physical well-being. By leveraging facial and voice recognition, the assistant detects emotional states in real time and offers personalized workout suggestions, mental health tips, and motivational content. Integrating Gemini’s API, it provides empathetic, conversational support that adapts to the user’s mood.

In-scope:

  • Emotional recognition via facial and voice analysis.
  • Personalized workout and mental health recommendations.
  • Integration with wearable devices for real-time feedback.
  • Gemini-powered empathetic interactions.
  • User profile creation and progress tracking.

Out of scope:

  • Direct medical or psychological diagnoses or therapy.
  • Long-term mental health treatments.
  • Advanced health data analysis beyond basic metrics like heart rate or steps.

Future opportunities:

  • Integration with advanced wearables to track biometric data (e.g., heart rate variability, sleep patterns).
  • Multilingual support for broader accessibility.
  • Partnerships with universities and wellness programs.
  • Collaborations with mental health professionals for more robust emotional support features.

Challenges we ran into

One of the key challenges we faced was accurately detecting emotional states through facial and voice recognition. Emotions can be subtle and vary across individuals, making it difficult for the AI to consistently classify them. We encountered a bug where the system often misinterpreted neutral expressions as negative emotions, skewing the recommendations.
To overcome this, we fine-tuned our models using a more diverse dataset, like using the Gemini API key for facial and vocal recognition and improved the emotion classification algorithm by incorporating additional facial features and audio cues. We also ran extensive tests with a variety of users to ensure the AI could better recognize a wide range of emotions, improving the system's accuracy and overall user experience.

Tracks Applied (1)

20. Gemini-Enhanced AI for Mental Health & Emotional Support

Our project tackles the challenge by using facial expressions and voice cues to determine how users feel. Based on that,...Read More

Discussion

Builders also viewed

See more projects on Devfolio