CEREBRO

Team Project for Undergrad Final Project (Most Innovative Project Award)
Spring 2014

“Real-time EEG based Object Recognition” is used to identify objects a person is thinking about and to display it on a screen for a person who is paralyzed or is not able to communicate in a feasible manner. A Brain Computer Interface (BCI) provides a communication path between human brain and the computer system. With the rapid advances in the areas of information technology and neurosciences there has been a surge of interest in turning fiction into reality. The major goal of BCI research is to develop a system that allows disabled people to communicate with other persons, to control artificial limbs, or to control their environment. The research area includes comparison of invasive and non-invasive technologies to measure brain activity, evaluation of control signals (i.e. patterns of brain activity that can be used for communication), development of algorithms for translation of brain signals into computer commands and the development of new BCI applications.



We presented a Research Paper


Data extraction

The EmoEngine communicates with the Emotiv headset, receives preprocessed EEG and gyroscope data, manages user-specific or application-specific settings, performs post-processing, and translates the Emotiv detection results into an easy-to-use structure called an EmoState. This step is used to obtain the electroencephalography (EEG) signals from the neuroheadset in the form of sensor readings and store it in a file. It interfaces with the device and obtains the required sensor readings, which can be subsequently processed.


Noise reduction

In this step, the idea is to reduce the noise in the EEG signals to some extent. The concepts of various mathematical and statistical methods will be used to reduce noise in these signals. Some advantages are: it reduces noise, lowers memory costs and algorithms become faster.


Classification

Finally, the filtered data is classified to one of the categories or classes established in the classification phase. The project employs the use a Machine Learning Algorithm to classify the data. Classification can be done using late learners or early learners approach. Late learners create the classification model only when a data element is being classified. The k-Nearest Neighbor algorithm falls into this category. Alternatively, the model can be made as soon as the training data is available. These early learner algorithms are used when the training set is relatively non-volatile. Decision trees and rule based classifiers are examples of this type



Gantt Chart for schedule






System Architecture






Data Flow Diagram






EEG Visualizations

A plot of EEG for trials (red, blue, green) for an apple fruit when the user saw an apple while wearing neuroheadset.









OTHER PROJECTS