3.4. Deep Learning Models

In this section, we dive into a wide array of deep learning architectures developed for emotion recognition—focusing less on performance and more on the diverse ways these models approach and handle EEG and ECG data. While previous section showed that simple machine learning models can produce decent results, this section highlights how advanced architectures can dig deeper into complex data, revealing hidden patterns. Each approach has its unique way of structuring data, from CNNs to transformers, offering interesting variations on the base models.