Engineer

Chowdhury Erfan Shourov

************@*****.*** erfanchowdhury.com Ashwood Ln San Jose, California +951******* EDUCATION

SAN DIEGO STATE UNIVERSITY

MSC IN ELECTRICAL ENGINEERING

GPA : 3.63/4.0

December 2020 San Diego,CA

NORTH SOUTH UNIVERSITY

BSC IN ELECTRICAL ENGINEERING

May 2016 Dhaka, Bangladesh

SKILLS

PROGRAMMING/FRAMEWORKS

Proficient In:

•Python 3 • C++ • Tensor Flow 2

•Keras • Scikit Learn • OpenCV

• Matlab • Octave • Git • LA

TEX

Familiarity With:

• HTML5 • JavaScript • CSS3 • MySQL

• Tableau • WordPress

CORE COMPETENCIES

•Hands On Experience with implementing a

diverse set of Machine Learning and Deep

Learning algorithms

•Well versed in solving Computer Vision

problems using Deep Learning

•Proficient in Algorithms and Data

Structures

•Competent in end-to-end machine learning

system design

RELEVANT COURSES

• Machine Learning

• Deep Learning

• Stochastic Signals and System

• Linear Control Theory

• Data Structures and Algorithms in C++

ALGORITHM-CENTRIC PROJECTS

INTERACTIVE PATHFINDER VISUALIZER

FRAMEWORK PYTHON/PYGAME

• User marks source and destination along

with obstacles

• Finds the path from starting point to

destination using :

• Dijkstra’s Algorithm

• A* Algorithm

• Breadth-First Search

• Depth-First Search

INTERACTIVE OBSTACLE AVOIDING

ROOM CLEANING ROBOT

FRAMEWORK PYTHON/ PYGAME

• User defines obstacles

• Robot cleans room using backtracking

algorithm

VISUALIZE SORTING ALGORITHMS ON

RGB IMAGES

FRAMEWORK PYTHON/OPENCV

•Visualization of

Selection,Insertion,Bubble,Heap,In-Place

Merge & Quick Sort algorithms on RGB

images

MACHINE LEARNING/ DEEP LEARNING PROJECTS

AUTOMATED BANGLA SUDOKU SOLVER USING TRANSFER LEARNING AND BACKTRACKING ALGORITHM

FRAMEWORK TENSORFLOW/KERAS/OPENCV/PYTHON

• Bangla is predominantly spoken by 98 % of Bangladesh’s population and serves as the nationa language of the nation.

• The first of its kind Sudoku Solver for this language

• Trained Bangla Digits using Transfer Learning with ResNet50

• Used hand-implemented backtracking algorithm to solve the 9 9 puzzle and process result in real-time

ROBUST REAL-TIME ROAD SIGN DETECTOR FOR AUTONOMOUS VEHICLES FRAMEWORK PYTHON/TENSORFLOW OBJECT DETECTION API

• Detecting Road Signs in Real-Time Using Single-Shot Multibox Detector

• Can detect ”Stop”,”Pedestrian Crossing” and ”Signal Ahead” Signs ROBUST REAL-TIME FACIAL EMOTION RECOGNITION SYSTEM FRAMEWORK PYTHON/KERAS

• Detecting Facial Emotion In Real-Time Using VGG Architecture

• Can detect six emotions : ”Neutral”,”Happy”,”Sad”,”Angry”,”Surprised” & ”Scared” EDVARDMUNCHANDVINCENTVANGOGHPAINTSSANDIEGOSTATEUNIVERSITY FRAMEWORK TENSORFLOW/KERAS

• Neural Style Transfer of San Diego State University with Edvard Munch’s Scream and Vincent Van Gogh’s ”Starry Night” with VGG19 pre-trained model using multiple GPUs.

THESIS

DETECTION AND CLASSIFICATION OF PEDESTRIANS AND SKATEBOARDERS FOR AUTOMATED CALCULATION OF POST ENCROACHMENT TIME AND EDGE-BASED AUTOMATIC HAZARDOUS CONFLICT ZONE DETECTION

FRAMEWORK TENSORFLOW OBJECT DETECTION API,OPENCV & GOOGLE CLOUD PLATFORM

April 2019 – September 2020

• Curated the first ever Skateboarder dataset to isolate pedestrian-skateboarder interaction using VGG Annotator

• Trained Faster R-CNN and Single Shot Multibox Detector (SSD) models with high accuracy to calculate surrogate safety measures.

• Deployed model on Google USB Accelerator and Raspberry Pi WORK EXPERIENCE

IOT LAB SDSU GRADUATE RESEARCH ASSISTANT

March 2019– November 2020 San Diego, CA

–Pioneered Detection and Classification of Pedestrians and Skateboarders for Automated Calculation of Post Encroachment Time and Edge-based Automatic Hazardous Conflict Zone Detection interdisciplinary Computer Vision project with Dr. Christopher Paolini using SOA Deep Learning techniques. RADAR AND LIDAR LAB RESEARCH ASSISTANT

January 2018 – December 2019 San Diego, CA

–Testing and Certifying Lidars and Radars for the California Highway Patrol,under the supervision of Dr. Lal Tummala

SAN DIEGO STATE UNIVERSITY TEACHING ASSISTANT

January 2019 – December 2020 San Diego, CA

–Mentored freshers in mastering basics of Programming Language AWARDS

2016 1st Prize Winner (USD $1200) of Idea Competition arranged by Ministry of Power, Energy and Mineral Resources (MPEMR), Bangladesh. Project : Security System and Maintenance for Transformers in the Rural Areas of Bangladesh PUBLICATIONS

[1] E. C. Shourov and C. Paolini. Laying the groundwork for automated computation of surrogate safety measures (ssm) for skateboarders and pedestrians using artificial intelligence. In 2020 Third International Conference on Artificial Intelligence for Industries

(AI4I), pages 19–22. IEEE, 2020.

Contact this candidate