Vishal K. Gupta is pursuing a Master of Science in Computer Science and Engineering at the University of Florida with a 4.0 GPA. He has a B.Tech in Electrical Engineering from Indian Institute of Technology Kanpur with a 7.0 GPA. He has over 3 years of professional experience as an Associate Engineer at Qualcomm India Pvt. Ltd. where he worked on software verification of security IPs and debugging firmware. He also has research experience working on fault detection, visual recognition, and condition monitoring of rotating machines using machine learning techniques. He is proficient in languages such as C/C++, Java, Python, and MATLAB.
I am a graduate student pursuing MS in Electrical Engineering (Embedded Systems) at the University of Colorado Boulder. Here is my resume which contains the list of projects I have completed so far, technologies that I have worked on, organizations I have been associated with and a patent that I have co-authored.
I am a graduate student pursuing MS in Electrical Engineering (Embedded Systems) at the University of Colorado Boulder. Here is my resume which contains the list of projects I have completed so far, technologies that I have worked on, organizations I have been associated with and a patent that I have co-authored.
Business Situation
Immumterix does analysis of DNA sequences and the immune system using multiple lab techniques. They needed to use computational techniques to analyze DNA through methods like high throughput analysis. They would analyze the DNA sequence and would represent each element through ASCII values. By studying the DNA sequence thoroughly, they were able to understand human DNA through various algorithms. What they needed was sequence alignment analysis and similarity searches in biological databases. Each sample would contain approximately 400 MB to 1 GB of data and would need customization based on the data collected. The need was to coordinate with the scientists from Immumetrix in order to understand the complexities and align the analytical algorithm accordingly. The objective of the project was to analyze the DNA data and provide outputs in CSV and graphical formats.
Business Situation
Immumterix does analysis of DNA sequences and the immune system using multiple lab techniques. They needed to use computational techniques to analyze DNA through methods like high throughput analysis. They would analyze the DNA sequence and would represent each element through ASCII values. By studying the DNA sequence thoroughly, they were able to understand human DNA through various algorithms. What they needed was sequence alignment analysis and similarity searches in biological databases. Each sample would contain approximately 400 MB to 1 GB of data and would need customization based on the data collected. The need was to coordinate with the scientists from Immumetrix in order to understand the complexities and align the analytical algorithm accordingly. The objective of the project was to analyze the DNA data and provide outputs in CSV and graphical formats.
I am a graduate student pursuing MS in Electrical Engineering (Embedded Systems) at the University of Colorado Boulder. Here is my resume which contains the list of projects I have completed so far, technologies that I have worked on, organizations I have been associated with and a patent that I have co-authored.
1. Vishal K. Gupta
(Master of Science, Computer Science and Engineering, University of Florida)
www.linkedin.com/in/vishalgupta4081
Email: vishal191@ufl.edu | vishal.gupta4081@gmail.com
Ph. No: +1 (352) 530-5982
Education
MS, Computer Science and Engineering ( Aug’16-Present )
University of Florida | GPA: 4.0/4.0
B.Tech, Electrical Engineering ( July’10-June’14 )
Indian Institute of Technology Kanpur, India | GPA: 7.0/10
Professional Experience
Qualcomm India Pvt. Ltd. | Associate Engineer ( June’14-Aug’16 )
Software centric verification of ARM TrustZone and virtualization extensions based IPs, drivers & Cryptographic engines
Debugged security software patches for successful development and integration of firmware on Snapdragon 820, 810, 617 etc.
Improvised backend verification infra by automation reducing man-hours by 3x & 2x boost in verification coverage
Qualcomm India Pvt. Ltd. | Interim Engineering Intern ( May’13-July’13 )
Detailed mathematical analysis of pre-silicon verification paradigm in which RTL design is verified against the design behavior
Performed formal verification of design FSM by cross multiplication against negate of property FSM
Debugged failing properties in AXI/AHB to QSB Protocol used in AXI cross-bar switch deployed across all Qualcomm SoCs
Intelligent Informatics Lab, IIT Kanpur in association with The BOEING Company | Research Associate (May’12-Dec’12)
Thoroughly analyzed acoustic signals in time, frequency and wavelet domain using MATLAB Signal Processing modules
Developed Softmax Classifier over features from Sparse Auto-Encoder network for vault prediction from acoustic emission
Prototyped the fault detection module on Android for increased versatility and validated its feasibility for Air Compressors
Skills
C/C++ (3+ yrs), Java (3+ yrs), Python (4+ yrs), MATLAB (5+ yrs), TensorFlow, RESTfull APIs, Spark, HTML, Linux, Android, LaTeX
Academic Projects and Research Experience
Transfer Learning on Kaggle Cats & Dogs Dataset ( Sept’16-Dec’16 )
Extracted CNN feature using TensorFlow API & Inception-v3 (GoogLeNet) framework pre-trained on ImageNet Dataset
Deployed Linear & Kernel SVM accompanied by hyper-parameter optimization using Bayesian approach
Registered 2-class accuracy as high as 62% with just 250 samples, 70% for 5000 samples and 63% for 3-class classification
Smart Sleep Assistant: IoT on XINU ( Sept’16-Dec’16 )
Designed a composite sensor “Sleep/Active” from thermostat & accelerometer data to monitor sleep conditions in real time
Deployed MQTT for edge/cloud communication, track of prior observations resulted in drastic decrease in energy consumption
Event-driven non-blocking server capable of handling 1000s of clients simultaneously for latency, performance and scalability
Mid-Layer CNN Features for Visual Recognition ( Feb’14-Apr’16 )
Exploited spatial feature representation which out-performed state-of-the-art methods with each added layer of spatial pyramid
Extended sample based sub-categorization approach to multi-class scenario for discovering unique intra-class discriminative
cues which further augmented classifier performance by 2-3%
Introduced spatial sub-categorization against sample based sub-categorization which outperformed the later by 4-5% on various
visual recognition task with SGD-SVM classifier
Relevant Coursework
Graduate: Advance Machine Learning • Sparse Coding in Sensing • Distributed OS Principles • Big Data Ecosystems • **Advance
Data Structures • **Analysis of Algorithms ( ** In progress )
Undergraduate: Machine Learning Techniques • Artificial Intelligence Programming • Computer Vision & Image Processing
Probability and Statistics • Natural Language Processing • Convex Optimization in SP/COM • Statistical Simulation & Data Analysis
Publication
Nishchal Verma, Vishal Gupta, Mayank Sharma & Rahul Kumar Sevakula, “Intelligent Condition Based Monitoring of Rotating
Machines using Sparse Auto-Encoders”, IEEE International Conference on Prognostic Health Management (IEEE PHM), 2013
Anup Zade & Vishal Gupta, “Software Centric Verification of xPU and other Security Blocks*” In QBUZZ’15, Annual
Conference organized by Qualcomm India Pvt. Ltd. (* Manuscript unavailable publicly due to Qualcomm’s Policy)