Shambhavi Shanker

Third Year Undergraduate

Email id: shambhavi13shanker@gmail.com
Contact: 9798192829
Electrical Engineering, Dual Degree(Btech + Mtech)
Indian Institute of Technology Bombay

Pursuing a Minor Degree in Machine Learning and Data Science, offered by CMInDS, IIT Bombay

S c h o l a s t i c Ac h i e v e m e n t s
• Secured an All India Percentile of 97.95 in JEE-Advanced among 0.15 million candidates (’21)
• Secured an All India percentile of 98.89 in JEE-Mains among 10 lakh candidates (’21)

I n t e r n s h i p s a n d R e s e a rc h E x p e r i e n c e
Unsupervised Domain Adaptation in Medical Imaging | MeDAL Lab, IIT Bombay (May’23-Ongoing)
Guide: Prof. Amit Sethi
• Read and implemented a paper on Attention-based deep Multiple-Instance Learning applied to the MNIST
dataset, and conducted experiments to evaluate its effectiveness when applied to Whole Slide Images(WSI)
• Experimenting with Domain Adaptation methods on WSI to leverage labeled data for obtaining labels for the unlabeled
dataset, reducing the cost & time required for unlabeled WSI in clinical practices and enhance performance across diverse domains
• Developing an algorithm utilizing domain adaptation techniques, inspired by research papers like Unsupervised Domain
Adaptation by Backpropagation (DANN, CDANN) and Target Versus Source Domain Discrepancy Minimization (TVT)
• Utilized attention-based multiple-instance learning techniques, and drawing inspiration from the CLAM approach, for the task
• Implemented Tissue Detection, Nuclei Counting, & Selection of highly-attended patches in WSI, to enhance classification accuracy
• Designed a custom Dataloader and Model Architecture incorporating ResNet50 using PyTorch to
enhance HPV detection performance & sensitivity in WSI and achieved accuracy of 76%
T e c h n i c a l P ro j e c t s
Camouflage Object Masking | Winter in Data Science | Analytics Club (Jan’23)
• Developed a robust camouflage object masking system using PyTorch framework and Computer Vision algorithms
• Learnt the basics of Deep Learning and Convolutional Neural Networks and reviewed research papers on Image
Segmentation, Feature Pyramid Networks for object detection and Camouflaged Object Segmentation
• Used COCO dataset to train and fine-tune a U-Net segmentation model to effectively detect & mask camouflaged objects
Sudoku Solver | Self Project (Jun’23)
• Designed an efficient Sudoku solver algorithm using Backtracking and Constraint Propagation techniques
• Utilized OpenCV for image preprocessing, including Noise Removal, Contour Detection, and Perspective Correction,
to accurately extract the digits from input Sudoku image, enhancing the efficiency and accuracy of the Sudoku solver
• Trained a neural network on the MNIST dataset to achieve 98.3% accurate recognition of the extracted digits from the image
Equilibrium Propagation for MNIST Classification | Neuromorphic Engineering (Nov’23)
Prof. Udayan Ganguly
• Implemented Equilibrium Propagation on a neural network model optimizing hyper-parameters including
Batch Size, Forward/Backward Pass Time, Time-step, and Clamping Factor for MNIST dataset tasks
• Demonstrated 90.84% accuracy and showcased the constant decrease in Hopfield energy, highlighting the
energy efficiency of neuromorphic computation over conventional Back-propagation
Autonomous Driving Vehicle | Summer of Code | Web and Coding Club (May’23-Ongoing)
• Implemented a Maze Solver by applying knowledge of Markov Decision Process(MDP), Value Iteration,
and Policy Iteration, utilizing these concepts to model the maze problem and find optimal paths efficiently
• Implemented the Windy Gridworld problem by acquiring knowledge in Temporal Difference Learning, Monte Carlo
methods, Q-learning, and SARSA, utilizing these algorithms to navigate and optimize paths in presence of wind disturbances
Design of Computing Systems | Course Project | Microprocessors and Digital Systems (May’23)
Prof. Virendra Singh
• Designed flowcharts, control logic and datapath for computing systems with their respective ISA’s provided
• Implemented a Multi-cycle processor, IITB-CPU for a 16-bit computer system, incorporating 8 registers and point-to-point
communication infrastructure utilizing VHDL’s capabilities to model the hardware components on Quartus
• Developed a 16-bit 6-staged Pipelined microprocessor, IITB-RISC-23 based on the Little Computer
Architecture, optimized for performance using hazard mitigation techniques like inter-stage data forwarding
• Developed a scaled-down version of the Intel 8085 microprocessor, Mini-8085, utilizing the hardware flow
chart method and a microcode-based architecture with a control store (CS) for storing encoded control signals
Lane Detection | Self Project (Dec’22)
• Developed a real-time lane detection system using Computer Vision techniques for autonomous cars
• Implemented algorithms for Lane Detection, including Canny Edge Detection, Hough Transforms, and
Image Segmentation and optimized its performance for real-time applications
Programming 8051 | Microprocessor Laboratory (Jan ’22-April’22)
Prof. Saravanan Vijayakumaran
• Developed embedded systems using Assembly and Embedded C to program Intel 8051-based Pt-51 microcontrollers.
• Implemented keyboard interfacing, LCD display, timers, and external interrupts for stopwatch and musical notes.
• Established serial communication using a USB-UART module and successfully executed a Lab Management System
• Used serial port interface (SPI) to interface an analog-to-digital converter (ADC) MCP3008 with the 8051 micro-controller

Simulation and Statistics | Probability and Random Processes (Aug ’22-Nov’22)

Prof. D. Manjunath
• Developed Python Algorithms to solve real world problems like Estimating the population average of
data consumption using different sampling methods, Maximising toss rewards using Hoeffding’s Lemma
• Implemented Linear Regression, Markov Process for data transmission, Capture-Release-Recapture problem,
Queueing problem leveraging Python libraries like Matplotlib, Seaborn, and Numpy for data analysis and visualization
Digital Logic Design in VHDL | Digital Circuits Lab (Aug ’22-Nov’22)
Prof. Maryam Baghini
• Created a String Detector using a Mealy type FSM which detects required sub-sequences in the input sequence of letters
• Designed a Moore type Finite State Machine (FSM) which acts as a 6-bit sequence generator in VHDL
• Created an Arithmetic Logic Unit capable of Additon, Subtraction, Comparison, and Multiplication of binary numbers
• Implemented the logic using Intel Quartus, deployed and tested in Xenon-10 FPGA board using Scanchain evaluation
XLR8 | Electronics and Robotics Club (Aug’22)
• Designed and tested a four-wheeled obstacle manoeuvring bot with differential steering, controlled by ESP32
• Designed bot’s chassis and electrical circuit to tune it for working in maximum Accuracy and Speed
• Used HTML web based interface to communicate with the robot via commands given on a mobile phone

Technical Skills
Languages C, C++, Embedded C, Python, VHDL, Verilog, Assembly, Simulink
Software Quartus, VS Code, ArduinoIDE, MATLAB, GNU Radio, NGSpice, JuPyter
Python Libraries Matplotlib, Pandas, Numpy, Tensorflow, Keras, OpenCV, OS, Pytorch
Publishing and Development Github, HTML, Excel, Git, LaTeX

C o u r s e s U n d e rta k e n
Electrical Microprocessors, Digital Systems, Control Systems, Analog Circuits, EM Waves,
Electronic Devices and Circuits, Signals and Systems, Communication Systems,
Digital Signal Processing, Probability & Random Processes*, Optimal Controls*
Computer Science Introduction to Machine Learning, Simulation Modelling and Analysis,
Digital Image Processing, Neuromorphic Engineering, Advanced Image Processing*
Physics Quantum Theory and its applications, Basics of Electricity and Magnetism
Mathematics Multi-Variable Calculus, Linear Algebra, Differential Equation, Complex Analysis
Others Molecular Cellular and Physical Biology, Engineering Drawing, Chemistry, Economics
Online Data Science Bootcamp conducted by WiDS,IIT Bombay
*(to be completed by Apr’23)
Positions of Responsibility
Media and Marketing Coordinator | Techfest’22
• Contacted 20+ Companies and 600+ POCs for association with Techfest during the event
• Successfully organized Virtual Ed-Conclave, an event with 10+ career guidance sessions by renowned
professors from various IITs with 40k+ student based viewership
E x t r ac u r r i c u l a r Ac t i v i t i e s
• Professionally trained in Bharatnatyam for 4 years and Kathak for 3 years and performed at intra and inter school events
• Participated and won several medals in Olympiads organized by SOF in school, and qualified for further levels
• Learnt to play Keyboard for 6 years and performed at cultural and inter-school events winning awards
• Participated in many inter-school Debate competitions, speech competitions and anchored at many cultural events