Project Members: Ferryn Gradney, Jacob Nguyen, Troy Lee, Ishaan Grewal, Devin Raye
Project Sponsor: William Trehern
The field of neuroergonomics has long aimed at using data from a multitude of brain scans and tests to improve and understand human lives. In the past, there has been work in classifying human emotions, using mental models to predict human stress, and improving brain functions via electrical stimulation. Recently, as with many similar fields, the focus has shifted towards heavy utilization of machine learning techniques and algorithms. Although past work has leaned slightly towards the use of CNN’s and SVM’s, as a relatively new and niche field, there are no set conventions or algorithms to use. Therefore, we plan to explore many different approaches in this project, using many different datasets.
For these past few weeks, our efforts have been directed towards research and familiarizing ourselves with the field. We’ve all been reading a neuroergonomics textbook covering past research, the types of input data, the types of neural networks used, and much more. We did this by assigning parts, taking notes, then presenting our notes to each other. This was important as none of us are familiar with this field and setting the groundwork is essential for future progress.
In addition, we have split into two groups, one focused on emotions and the other on fatigue. This was in preparation for our collaboration with two professors.
For the emotion group, we have been going over the DEAP dataset, which consists ofan online assessment for volunteers to rate music videos based on arousal, valence and dominance; and participant ratings, EEG and physiological recordings and face videos where volunteers watched a subset of the music videos rated in the online assessment.
Meanwhile, for the fatigue group, we’ve been busying ourselves with reading Dr. Mehta’s previous publications. She’s done a lot of work in the fields of both mental and physical fatigue. For example, she profiled fatigue’s signature in people of various conditions, quantified the effects of fatigue on mental capacity, and much more. Accordingly, we’ve been narrowing down the possible datasets with the aim of picking a single dataset to start with by Saturday 10/31.
A recurring problem we have encountered during the duration of this project is scheduling time for all the team members to meet and the effects that come with there being no possibility for us to meet in person. The members of our project team have used applications such as discord and email to communicate, but it is more difficult to keep track of each other’s progress and to collaborate with one another as we are restricted to working as a team online. There is also a learning curve as the team members have varying levels of experience in machine learning, neuroergonomics, and data sorting. As the project progresses, we hope to have each team member get caught up enough to be on the same level of understanding.
In the following weeks, we will be able to extract, understand, plot and visualize the data in the datasets chosen. In addition to that, we will be going over more topics and applications of the field of neuroergonomics using the same process as mentioned in the “Current Progress.” Once we have a better general understanding of neuroergonomics, we will then start to build a deeper interpretation on how neuroergonomics can be used in a professional setting.
Within the next month, we will have a greater comprehension of the data through data augmentation, sampling and model fitting in order to collaborate with Dr. Lench in the Psychology Department. With the collaboration, we will further expand our knowledge on EEG data and how to apply it to the project by understanding how emotions and mental states are quantified to be processed by computer interfaces. In doing so, we will be able to potentially alter an individual’s emotional and mental state.
After working with Dr.Lench, we plan to work with Dr. Mehta in the Industrial Engineering Department to continue our research and application of neuroergonomics. We will use the ideas of neuroergonomics to help eliminate wastefulness in production processes and create more efficient systems for workers to complete a task given to them. Since neuroergonomics is the study of the performance of the human brain in a professional and everyday setting, we hope to be able to help increase efficiency in both settings.
Through this project, we hope to expand the possible applications of the relatively new field of study in any environment or setting to further interpret behavioral performance, and to enhance performance by combining neuroscience, the study of the brain structure and function, and ergonomics, the study of matching technological advances with the capabilities of people as to create a more productive and safe work environment.