-
-
camera page
-
home page
-
home page
Inspiration
The inspiration for EEPY came from the need to enhance driver safety by detecting signs of fatigue. We realised that this technology could also be applied to other areas where monitoring alertness is crucial, such as in workplaces or during long study sessions.
What it does
EEPY is an AI-powered fatigue detection system that monitors eye and mouth movements to detect signs of drowsiness. It uses a camera to alert users when they show signs of fatigue, helping to prevent accidents and improve productivity. We also implemented an AI chatbot that is activated by the user and checks in on their fatigue and wellbeing levels.
How we built it
We built EEPY using a combination of computer vision and machine learning techniques. The system uses OpenCV and dlib to detect facial landmarks and calculate Eye Aspect Ratio (EAR) and Mouth Aspect Ratio (MAR). These metrics are used to determine if the user is showing signs of drowsiness. The data is then sent to a web interface using Flask and Socket.IO, where it is displayed in real-time. The AI chatbot in EEPY is implemented using the OpenAI API and LangChain framework. It leverages the ChatOpenAI model to generate responses based on user queries. The chatbot uses tools like DuckDuckGo and Wikipedia for information retrieval. Speech-to-text and text-to-speech functionalities are handled by speech recognition and pygame libraries. The chatbot listens for activation words, processes user queries, and provides concise responses, enhancing user interaction.
Challenges we ran into
One of the main challenges we faced was accurately detecting facial landmarks in various lighting conditions and angles. We also had to fine-tune the thresholds for EAR and MAR to minimise false positives and negatives. Integrating the real-time data streaming with the web interface was another challenge that required careful synchronisation.
Accomplishments that we're proud of
We are proud of successfully creating a system that can accurately detect signs of fatigue and alert users in real-time. The integration of various technologies, from computer vision to web development, was a significant achievement. We are also proud of the user-friendly interface that makes it easy for users to monitor their alertness.
What we learned
Throughout this project, we learned a lot about computer vision, machine learning, and real-time data streaming. We also gained experience in integrating different technologies and working with APIs. The project taught us the importance of thorough testing and fine-tuning to achieve accurate results.
What's next for Eepy
In the future, we plan to improve the accuracy of the fatigue detection system by incorporating more advanced machine learning models. We also aim to expand the application of EEPY to other areas, such as monitoring alertness in workplaces and during long study sessions. Additionally, we plan to add more features, such as emotion detection, to enhance the effectiveness of drowsiness detection.
Log in or sign up for Devpost to join the conversation.