Welcome to the Cognitive State Analysis repository! Here, we explore the fascinating world of cognitive states using cutting-edge deep learning models: EEGNet and ATCNet.
Understanding cognitive states is like deciphering the brain's secret code. Here are the key states we'll explore:
-
Attention π: Imagine your brain as a spotlight. When you're focused on a task or a captivating movie, your attention state is in full swing. EEG helps us peek into this spotlight.
-
Relaxation π: Ah, the blissful moments when stress melts away. Relaxation is essential for mental well-being. EEG captures those serene brain waves during meditation or a lazy Sunday afternoon.
-
Alertness π¨: Ever had that "I need coffee" feeling? Alertness keeps us awake, vigilant, and ready to tackle challenges. EEG reveals the brain's wake-up call.
-
Meditation π§ββοΈ: Picture a serene mountaintop. Meditation takes us there. EEG shows us the brain's zen mode.
-
EEG: Electroencephalogram records brain activity using electrodes placed on your scalp. It's like eavesdropping on neurons having a chat. π§ π¬
-
EEG Data: These brainwave recordings come in time-series form. Think of them as brain symphoniesβeach electrode playing its unique melody. π΅π
- Description: EEGNet is our brain-savvy neural network.
- Features:
- π Depthwise separable convolutions: Efficient feature extraction.
- β±οΈ Temporal convolutional layers: Capturing brain rhythms over time.
- ποΈ Lightweight design: Perfect for real-time applications.
- Description: ATCNet combines attention and convolutions for brainwave magic.
- Features:
- π Multi-head self-attention: Like brain detectives focusing on crucial moments.
- π Temporal convolutions: Surfing the brainwaves for patterns.
- π State-of-the-art performance: ATCNet knows its cognitive ABCs.
- Complex networks and deep learning for EEG signal analysis: https://link.springer.com/article/10.1007/s11571-020-09626-1
- Research Paper: https://www.mdpi.com/2306-5729/4/1/14
- MSHANet: a multi-scale residual network with hybrid attention: https://link.springer.com/article/10.1007/s11571-024-10127-8
- EEG Motor Imagery Deep Learning Repository:https://github.com/edw4rdyao/eeg_mi_dl
- EEG classification using deep 1D convolutional neural network: https://youtu.be/qLYVnB4pWI8?si=99y4Q3aScFfRJ41U
Feel free to fork this repository. If you need more brainpower , just ask! π§ π€π