Join us for December’s Research Café, where we’ll explore how creative and scientific innovation intersect from graffiti and street art in the algorithmic economy to machine learning models that help us better understand brain white matter and traumatic injury. This virtual session will feature presentations by Ph.D. students Luis Contreras (Communication, Information, and Media) and Parameshwaran Pasupathy (Mechanical and Aerospace Engineering), followed by time for questions and discussion with the audience.
Research Café is one of SGS’s signature community-building events, bringing graduate students together each month to share ideas, learn from one another, and celebrate the breadth of scholarship across Rutgers. Learn more about our presenters and their research below.
Luis Contreras is a Ph.D. student in Communication, Information, and Media at Rutgers University’s School of Communication and Information. His research examines how graffiti and street artists use digital technologies and platforms to navigate visibility, resistance, and creative labor in the algorithmic economy. Drawing from digital ethnography and critical platform studies, his work explores the intersections of art, activism, and media infrastructures. He has presented his research at national and international conferences and teaches courses in public speaking and media production. Outside of academia, he supports community-based arts and storytelling initiatives that bridge culture, technology, and public engagement.
Bombing The Information Superhighway: Graffiti, Platform Infrastructures, and the Visual Politics of Algorithmic Culture
My research examines how graffiti and street artists adapt their analog and urban practices to the digital platform economy. Through qualitative interviews and digital ethnography, I analyze how artists navigate algorithmically driven platforms such as Instagram, integrating tools like drones, augmented reality, and virtual reality to extend graffiti’s visibility and resistance into digital spaces. The study conceptualizes this adaptation as a new form of “bombing,” where artists seek hyper-visibility across both physical and platformed environments. I explore how these practices involve new kinds of digital labor—coding, drone piloting, and video editing—while interrogating tensions between authenticity, commercialization, and algorithmic control.
Drawing from critical platform studies, infrastructure theory, and visual communication, I frame graffiti as both a symbolic act of resistance and a form of digital labor. The project also addresses how artists confront the vulnerabilities of hypervisibility, including surveillance, cultural appropriation, and the loss of anonymity. By tracing graffiti’s evolution through emerging media technologies, my research reveals how this subculture both leverages and resists the infrastructures of digital capitalism. Ultimately, the study contributes to understanding how creative subcultures repurpose digital infrastructures to assert presence, political expression, and autonomy within algorithmically mediated publics.
Parameshwaran Pasupathy is a PhD candidate in the Department of Mechanical and Aerospace Engineering at Rutgers University. Paramesh’s doctoral research is on the multiscale computational characterization of the mechanics associated with traumatic injury to the central nervous system, brain white matter (BWM). His research focuses on developing computational models and scientific computing software that rely on interdisciplinary areas such as computational mechanics, uncertainty quantification, non-local peridynamic formulations, and high-performance computing (HPC). Paramesh currently interns at Dassault Systèmes as a Cardiovascular Biomechanics engineer, making significant contributions to the development of the Living Heart Project, a high-fidelity multiphysics model of a healthy, 4-chamber adult human heart and proximal vasculature using high-performance computing and machine learning. Paramesh has a master’s degree in aerospace engineering from the University of Michigan at Ann Arbor.
A Machine Learning-based Constitutive Modeling of Fractional Viscoelasticity in Brain White Matter
Predicting the mechanical response of brain white matter (BWM), even in the limit of small strains is challenging owing to the inherent anisotropy of the three-dimensional microstructure and the various interactions between the structural components of brain tissue. Conventional viscoelastic characterization of BWM, typically carried out within the classical framework of springs and dashpots expressed as a Prony series, remains a purely empirical representation that is difficult to physically interpret.
MRE measurements increasingly suggest that the mechanical response of BWM exhibits power-law behavior. A power-law model in the frequency domain, under the assumptions of linear viscoelasticity and causality, yields a fractional viscoelastic model in the time domain. While fractional viscoelasticity provides a succinct, physically grounded framework well-suited for modeling and quantifying the power-law behavior observed in BWM, its numerical implementation is computationally expensive.
Here, we aim to develop a surrogate deep learning model that replaces the finite element (FE) model for characterizing BWM mechanical response under quasi-static loading. The forward problem of predicting the evolution of stress for a given geometry—typically performed using FE modeling—is replaced by a recurrent neural network (RNN) that learns the structure–property relationship between BWM microstructure, material properties and stress-strain behavior. The trained RNN emulates the forward FEM model, significantly accelerating the prediction of the stress-strain response of the microstructure. When integrated into an optimization workflow for estimating homogenized BWM properties, the RNN permits rapid evaluation of forward predictions at a fraction of the computational cost, enabling near–real-time estimation of BWM material properties.