Advancement: N. Nasiri - Extracting Implicit Features from Touch

Description: My dissertation proposal aims to extract implicit features from touch's cognitive and affective content through a tangible device equipped with pressure sensors. By utilizing machine learning, I seek to identify subtle cues and sensory information conveyed through touch and develop a high-dimensional representation of affective and cognitive states. My research aims to relate tactile content to cognitive and affective self-regulation tasks undertaken by individuals with ADHD as part of an NIH study. I also plan to train a large language model to identify the meaning of cognitive and emotional sentences in an embedded space. Finally, I want to capture touch's emotional and cognitive content in this high-dimensional form.

The proposed work seeks to extract meaning from an under-studied input modality (touch) and improve the fidelity of low-dimensional valence-value models that commonly underly efforts in affective computing. If the technology proves to be effective, it has the potential to facilitate numerous applications that utilize touch to convey emotions between humans and machines. This could include developing a therapeutic device that helps individuals with ADHD regulate their behavior and applications that respond to a user's cognitive and emotional state based on how they interact with their mobile devices through touch.

Event Host: Nahid Nasiri, Ph.D. Student, Electrical & Computer Engineering

Advisor: Daniel Shapiro and Gabriel Elkaim 

Dial-In Information

Zoom -

Passcode: 437596

Thursday, May 18, 2023 at 8:00am

Virtual Event

Recent Activity