Emotion detection is often presented without proper evaluation or deployment considerations. This project focuses on building a clear, testable emotion classification pipeline suitable for real-world integration.
Input Data
→ Preprocessing
→ Feature Extraction
→ Classification Model
→ Emotion Output
The pipeline is modular to allow experimentation with different models and modalities.
- Modular pipeline for easy experimentation
- Emphasis on data preprocessing
- Simple baseline models before complexity
- Clear separation of training and inference
- Emotion classification
- Preprocessing pipeline
- Model evaluation
- Predictive inference
pip install -r requirements.txt
python main.py