Data Labeler and Model Trainer for Emotion-Based AI Music Player
Developed a convolutional neural network (CNN) trained on over 3,500 labeled facial images for emotion classification. Managed the data preprocessing and labeling phases to ensure high-quality training data for emotion recognition tasks. Integrated labeled data to improve emotion detection accuracy across seven emotion states. • Labeled images were annotated for emotion categories such as happy, sad, angry, surprised, and more. • Conducted quality control and data cleaning to maximize annotation accuracy. • Used labeled data to train machine learning models achieving approximately 82% emotion detection accuracy. • Ensured dataset diversity to reduce model bias and improve generalization.