😲 Your Face is the Playlist: Building an Emotion-Aware Android App

The EmotionToMusic-App is an open-source Android project that uses computer vision and machine learning to recommend music based on a user's emotions. The app analyzes the user's face in real-time, using a camera to detect their emotions and play music accordingly. The app is built natively in Kotlin and follows modern Android development practices, with a core philosophy of on-device inference to ensure privacy, low latency, and offline capability. The tech stack includes CameraX, TensorFlow Lite, and Google ML Kit, which work together to detect faces, classify emotions, and play music. The app functions as a continuous loop, analyzing video frames frame-by-frame to detect the user's emotions and play music that matches their mood. The implementation involves several components, including the Analyzer, which processes frames and detects faces, the Model, which interprets the face and outputs probabilities, and the Mapper, which maps the result to a playlist. The app also includes design decisions and challenges, such as handling low light conditions and smoothing out emotion detection to prevent jitter. Overall, the EmotionToMusic-App is a great example of how accessible AI has become, and how developers can build smart apps by wiring together different components. The app's code is available on GitHub, providing a valuable resource for developers who want to explore the intersection of computer vision and music recommendation. The EmotionToMusic-App has the potential to revolutionize the way we interact with music, making it a more personal and emotional experience.