Mood Master is a website that helps people with mental disorders learn more about their actions and emotions. It has three main tools to assist them including a video emotion software, chatbot, and text sentiment analysis. Mood Master has a full encrypted login system implemented with Firebase.
My motivation for creating Mood Master was to help people with bipolar disorder learn more about their actions and emotions, and to use the power of technology to make this tool more accessible and engaging for the users.
Mood View is a feature that uses a video software to capture and analyze the facial expressions of the user in real time or with a video/image upload. It uses a facial detection algorithm to detect the user’s face and a facial expression classification algorithm to classify the user’s facial expression.
The video above shows the MoodView feature in action. It captures the user’s facial expressions and displays the results in real time.
MoodBot is an interactive chatbot that can provide personalized feedback and suggestions to the user. It uses the results from Mood View to help the user learn more about their actions and emotions. Implements automated vector caching. Powered by OpenAI.
The MoodBot chat interface.
Mood Text is a feature that uses a message analysis tool to detect the tone and sentiment of the user’s texts and offer guidance and support. Uses a fine-tuned transformer model to detect.
Mood Text sentiment analysis in action.
Mood Master has a full encrypted login system implemented with Firebase. The user can sign up and log in to the website using their email and password. Fernet encryption is applied to withstand potential attacks.
The Mood Master login page.
I learned a lot from building this project, such as how to use facial detection and emotion classification algorithms, how to use message analysis tools, how to design a user-friendly and interactive interface, how to optimize the performance and efficiency of the website, and how to publish the website on the internet.