Ph. D. Students Colloquium

Xixi Deng

Title: 
Reconstructing the Physical Properties of Translucent Objects: Applications in Art, Science, and Beyond

Abstract: 
Reconstructing the real world is a crucial yet challenging problem for applications in both entertainment and science. To address this challenge, we study inverse problems in transport theory, including radiative transfer and neutron transport. In these problems, the field (flow of energy) is described by transport equations. Our key challenge lies in determining the parameters of these equations from measured radiation data, such as transmittance and reflectance. This work has applications across multiple disciplines. I will present our research on reconstructing the geometry and optical properties of semi-translucent and thin objects from photographs, and discuss our recent advances in using Monte Carlo simulation gradients to optimize nuclear reactor shielding design.

Bio:
Xi is a Ph.D. candidate in the Graphics and Vision Group at Cornell University, advised by Steve Marschner. Her research interests are physically-based light transport simulation, inverse rendering, and Monte Carlo sampling. Before she started her Ph.D. at Cornell (June of 2019), she was a master's student in the Digital Arts program at Dartmouth College. She did her master’s thesis in the Visual Computing Lab under the supervision of Wojciech Jarosz. She has been awarded an Adobe Research Fellowship, Booking.com Research Fellowship, and was an NVIDIA Research Fellowship Finalist.

Gloire Rubambiza

Title: 
Towards Resilient Rural Networking Infrastructure: Seamless Visions, Seamful Realities

Abstract: 
Data is transforming virtually every industry in the 21st century. The amount of new data generated by IoT devices is expected to grow from 33 Zettabytes (ZB) in 2018 to 175ZB in 2025. However, existing networking infrastructure cannot keep pace with the IoT data, especially in US rural areas. This work examines how to build resilient rural networked systems for agricultural IoT applications in this Zettabyte era. In this talk, I will briefly discuss how we do so by combining networking and HCI. 

First, we create rural networked systems that are vendor agnostic, resiliently allowing IoT data transmission, collection, and analysis, while autonomously detecting and repairing failures. We present the Software-Defined Farm (SDF), a cloud-based architecture that advances the state of the art towards resilient IoT systems. Critically, we upscale our technical systems investigation with HCI by using the SDF system building to analyze how researchers envision and orient toward the eventual use of such systems in rural areas. Our analysis shows a consequential contrast between researchers’ seamless visions of rural networked system deployments and the seamful realities of their own system building.

In concluding, we show that our integrated approach of technical and sociotechnical aspects is one of the core contributions from my dissertation. Specifically, we demonstrate that the networking and HCI advancements presented together in this work would not be possible without the integrated approach. We call the approach trilingualism, which entails the integration of deep technical work, its domain applications, and critical reflections on the technical decisions. Together, trilingualism provides a step in advancing the state of the art in networking for the Zettabyte era. I wrap up the talk with some future research directions on applying trilingualism to build resilient networked systems in space.

Bio:
Gloire Rubambiza is a recent CS PhD alum at Cornell University, where he conducted research in resilient networked systems with an emphasis on their societal impacts. Specifically, he implements, deploys, and anticipates the societal impact of networked systems for agricultural applications on rural US farms. As a Cornell student, he was a University Fellow and a fellow of NSF National Research Traineeship in Digital Plant Science. He regularly publishes in ACM, USENIX, and IEEE conferences. Beyond research, Gloire is passionate about broadening participation in computing, for which he recently won the Cornell Bowers CIS Distinguished Leadership in Service Award, a SUNY Provost Diversity Fellowship, and the Best Doctoral Presentation in the Doctoral Consortium at the 2023 ACM Richard A. Tapia Celebration of Diversity in Computing Conference. In his free time, Gloire enjoys soccer, travel, and cooking. Since leaving Cornell, he has been leading engineering efforts at Satlyt, a startup on a mission to interlink satellites into a virtual datacenter for space edge computing services.