Hi, my name is Vibhor Gupta.
Hi, my name is Vibhor Gupta. I am currently pursuing my Masters Degree in Electrical Engineering and Computer Science (EECS) from the University of Michigan.
Over the years, I have honed my expertise in a variety of areas, including Communication Networks, Wireless Systems, Reinforcement Learning and Real-Time Embedded Systems. Throughout my career, I have had the privilege of working with industry leaders such as Qualcomm and Samsung, where I played integral roles in the development and optimization of cutting-edge technologies.
In addition to my technical prowess, I am a dedicated leader and community advocate. As the Projects Officer for the IEEE HKN Chapter at the University of Michigan, I organized and facilitated community service opportunities, fostering a culture of giving back and collaboration. My leadership experiences extend to mentoring teams and coordinating events.
I am driven by a relentless pursuit of excellence and a commitment to making a positive impact in the world of technology. Let's connect to explore opportunities for collaboration, innovation, and growth.
Work Experience
University Of Michigan, MI, USA
Research Assistant, Guided by Prof. Lei Ying
Aug 2023 - Apr 2024
Working on development of a Reinforcement Learning algorithm to handle scheduling and queue length management under heavy traffic conditions in Wired and Wireless Communication Systems.
Qualcomm, CA, USA
mmW Modem SW Intern
May 2023 - Aug 2023
Redesigned and implemented an algorithm in a Real Time Embedded System, that allows for a faster convergence of Residual Side Bands in a direct conversion receiver.
Qualcomm, Hyderabad, IN
V2X RFSW Engineer
Jan 2020 - Aug 2022
Spearheading the design and commercialization of the first DSDA+V2X modem.
Samsung Research, Noida, IN
Network Engineer
June 2019 - Jan 2020
Developed Modem protocol procedures for support in Galaxy A Devices.
Research Projects
RL for NLP Tasks
University Of Michigan, MI, USA
Jan 2024 - Apr 2024
This project investigates methodologies for enhancing the performance of small language models by aligning them with user intent across diverse tasks with Reinforcement Learning from Human Feedback (RLHF). By leveraging RLHF, the aim is to address the limitations of smaller models, improve their coherence, consistency, and relevance in generating text, and mitigate biases inherited from training data. Specifically, the project presents a two enhancements for a parameterized masking strategy, that focuses model attention on relevant tokens, generating relevant coherent factually correct response.
REDD - REAL-TIME EXPEDITED DISEASE DETECTION
University Of Michigan, MI, USA
Aug 2023 - Dec 2023
In this paper, we present REDD: a Real-time Expedited Disease Detection application for Low Resource Countries and settings. Our system allows doctors and health care professionals to simply upload an image of chest x-ray scan via a Rasberry Pi4 and receive a segmented output scan of the input image with a diagnosis of the detected thoracic disease with up to 80% confidence score with DICE.
TCP in mmW Networks
University Of Michigan, MI, USA
Jan 2023 - Apr 2023
Implemented a new TCP algorithm for mm-wave networks, carried extensive analysis, investigated the claims and suggested further improvements for current TCP algorithms catering to the requirements of mm-wave networks.
Cognitive Radios in 5G
TU Wien, Vienna, Austria
May 2018 - Aug 2018
Increasing Spectrum Efficiency by Channel Detection,an implementation for COGNITIVE RADIO in 5G Networks.