Reconnect: Rehabilitation Online

HOME PAGE
Live Video Chat with Zen-AI Demo (Both audio and video alone are supported)
Community Chat Page

TEAM NAME: 4d Devs

TEAM MEMBERS:

Dhanush B – Developer and Designer
Dhanush U – ML Engineer
Dhanush K Gowda – ML Engineer
Dhanush N Prabhu – Developer

CONTACT EMAIL:

talktodhanush4554@gmail.com

PROJECT TITLE:

RECONNECT: Rehabilitation Online

HACKATHON THEME:

AI for Societal Good

PROBLEM STATEMENT:

Build an AI-powered chatbot or virtual assistant that leverages Gemini's natural language understanding and empathy to provide mental health and emotional support to students.

MOTTO / VISION:

"To build a personal AI counselor that's always with you, offering real-time emotional understanding, privacy, and secure support."

Inspiration

Our inspiration comes from the growing need for mental health and emotional support, especially for people who struggle with addiction, depression, and anxiety. Often, individuals face stigma, privacy concerns, and lack of immediate help, which can prevent them from seeking support. We wanted to create a solution that is always available, non-judgmental, and offers real-time, AI-powered emotional care.

What it does

Reconnect is an AI-powered personal counselor that provides real-time emotional support. It uses advanced AI technologies to detect emotions through voice, facial expressions, and text, offering personalized responses based on emotional cues. The platform also fosters a sense of community through shared experiences and success stories, ensuring users don’t feel alone in their journey to recovery.

How we built it

We built Reconnect using a combination of modern web technologies like React.js, Flask, Node.js, and MongoDB for the platform, and integrated AI/ML models such as wav2vec2 and FaceAPI for real-time emotion detection. Vertex AI and Eleven Labs powered our natural language understanding and voice processing. WebRTC was employed for real-time video analysis to capture emotional data.

Challenges we ran into

One of the main challenges was ensuring real-time emotion detection with low latency while processing audio and video data. We also faced difficulties with accurately interpreting diverse emotional cues due to cultural differences, which required extensive training of our AI models. Moreover, ensuring ethical AI behavior and safeguarding user data privacy were critical aspects that demanded special attention.

Accomplishments that we're proud of

We’re proud to have created an AI that can provide empathetic, real-time support to users based on their emotional state. Successfully integrating emotion detection from multiple channels (voice, facial expressions, and text) was a significant technical accomplishment. Additionally, the community support feature we implemented adds a human touch, providing users with stories of triumph to inspire hope.

What we learned

We learned the importance of balancing technology with empathy in AI-driven projects. Understanding the nuances of emotional detection, ensuring real-time capabilities, and maintaining ethical standards in AI development were crucial lessons throughout this journey. We also gained a deeper understanding of the complexities involved in ensuring privacy and security while dealing with sensitive user data.

What's next for Reconnect: Rehabilitation Online

Our next steps include improving the accuracy of our emotion detection models by training them with more diverse datasets. We also plan to introduce multilingual support and enhance the personalization aspect of the platform by adding AI-driven recommendations for mental health resources. Expanding the community feature to include real-time peer support and continuing to scale the platform for a global audience are on our roadmap.

CONTACT DETAILS
Email: talktodhanush4554@gmail.com
Ph: +91 8618541131
LinkedIn: Dhanush B