FURI | Summer 2020

A Benchmarking Framework for Data-Driven Compressive Sensing

Data icon, disabled. Four grey bars arranged like a vertical bar chart.

The objective of the research is to develop a benchmarking framework with a unified API to allow researchers to quantitatively evaluate new compressive sensing algorithms.To achieve fair and extensive comparisons,all data-driven methods are implemented in PyTorch,and are compared to state-of-the-art model-based methods with a variety of dataset and parameter setups.The CSGAN algorithm was reimplemented in PyTorch as part of this framework,based on an existing TensorFlow implementation.However,it was found that the original implementation of the research paper has been missing key components,and further investigation is required to reproduce the results from the original CSGAN research paper.

Student researcher

Márk Lakatos-Tóth

Computer science

Hometown: Budapest, Pest megye, Hungary

Graduation date: Fall 2020

More projects from the current symposium

Ana Girish

Aerospace engineering

Fabrication of Anisotropic Resin Polarizers with Liquid Crystal via Vat Photopolymerization for Enhanced Optical Steganography

Optimizing liquid crystal resin polarizers for optical steganography improves message security and accessibility, advancing data privacy.

Mentor: Xiangjia Li

Program: FURI

Nguyen Phuong Thao Tran

Chemical engineering

Urea formaldehyde polymerization

The findings have the potential to reduce carbon emissions, global warming and create new materials for the construction.

Mentor: Dorsa Parviz

Program: FURI

Minju Yoo

Mechanical engineering

4D Printed Structures with Tunable Surface Roughness based on Hygro-responsive Polymer and Vat Photopolymerization

Design and fabricate a structure capable of dynamically interacting with its surroundings through adjustable surface roughness.

Mentor: Xiangjia Li

Program: FURI