Designed Convolutional Neural Network model optimization strategies for Image Super Resolution and Denoising applications
Designed Neural Architecture Search methods for Mixed Precision Quantization and efficient hardware-aware neural networks for Graphics related problems using Unet model
Worked in the MIGraphX team (May 2019 - August 2019)
Developed Post Training Quantization (PTQ) methods to reduce the CNN weights from FP32 to Int8 precision.
Implemented the quantization algorithms and analysis on Vgg16, ResNet50, InceptionV3, Xception benchmark, resulting in negligible accuracy loss on Imagenet
Research Centre Imarat, Defence R&D Organization, Hyderabad, India
Undergraduate Technical Intern (May 2016 - June 2016)
Bharat Dynamics Limited (BDL), Hyderabad, India
Undergraduate Technical Intern (December 2015)
Academic Experience
Graduate Research Assistant
Supervisor: Dr. Arun K. Somani
Graduate Teaching Assistant
Teaching Assistant for the following courses:
Digital Logic Design (Undergraduate Course) for Fall 2017 and Spring 2018
Fault Tolerant Computing Systems (Graduate Course) for Spring 2020 and Spring 2022
Responsibilities included:
Conducting weekly lab sessions to assist students in writing code using Verilog HDL.
Guiding students on their final course projects, specifically focused on deploying applications on FPGA.
Holding office hours to provide one-on-one support and address students' queries effectively.
Evaluating homeworks, lab reports, and exams, while offering constructive feedback to enhance learning outcomes.
Graduate Courses
Machine Learning/Stats: Advance Design and Analysis of Algorithms, Deep Learning: Theory and Practice, Machine Learning, Probabilistic Methods, Statistical Methods for Machine Learning, Statistics Theory for Research
Systems: Applications of Parallel Computers (CS267- Berkeley; CprE 594 at ISU), Computer System Architecture, Fault Tolerant Computing Systems, High Performance Communication Networks, Real Time Systems
Research Award by Graduate and Professional Student Senate (GPSS) society at Iowa State University,
Spring 2023 [Certificate]
Selected for Oxford Machine Learning Summer school 2022 (OxML) in ML for Health and ML for Finance tracks [Certificate] (Acceptance Rate < 10%)
Our survey paper “Neural Architecture Search Survey: A Hardware Perspective,” has been identified as one
of the must-read AI papers in 2022 by a group of industry experts [URL]
