Arup Kanti Dey

Data Scientist, Data Analyst, Machine Learning, Deep Learning & Computer Vision Engineer| Delivering 94% accuracy through Computer Vision Techniques

SCoRe Lab, University of South Florida 5.1 Yrs , BroTecs Technologies Ltd. 2.2 Yrs

University of South Florida Phd Computer Science, University of South Florida Master of Science (MS) Computer Science, Military Institute of Science and Technology Bachelor of Science (BSc) Computer Science and Engineering

Data Scientist 02.00 Yrs, Machine Learning Engineer 02.00 Yrs, Data & Machine Learning Scientist 02.00 Yrs,

ython,  Machine Learning,  Java

About Arup Kanti Dey

I worked as an iOS app developer for 2 years before I embarked on my journey to get a PhD under the supervision of Prof. Dr. Sriram Chellappan at the Social Computing and Research Lab, University of South Florida. My experience of working at this lab exposed me to the power of Data Analytics and Machine Learning Algorithms. I got the opportunity to work on some excellent projects focusing on Healthcare and Social Welfare.

I got to work on an App-based Personalized Post-Cardiac Rehabilitation System. The project not only invited a grant of $25000 from the USF Bull Ring Accelerator Grant Program, but also $100,000 USF Strategic Investment Pool (SIP) Awards. Or take for instance, the Detecting Distracted Driving from Images project which detects distracted drivers with a staggering accuracy of 94%. The outcome of this study has also been published in many reputed academic journals. There have been many others like these. For all these projects, I have handled the various stages of the cycle such as gathering user requirements, application development, AI implementation, system troubleshooting and even managing users’ feedback.

Backed by more than a half decade of Industry as well as Research experience, I could be your go-to person for Deep Learning, Computer Vision Techniques, Machine Learning.

Technical Skills:

- Language: Python, Java, C++, C, R, Objective-C
- Database: MySQL, NoSQL, Oracle
- Technology | Platform: Android development, iOS development, AWS, Tableau, TensorFlow, MATLAB, ANACONDA
- Libraries: Scikit-Learn, Scipy, NumPy, Pandas, Matplotlib, Seaborn, OpenCV, Keras, imgaug, PyTorch

Total Work Experience: 04.00 Yrs

Domain Experience : Aerospace domain,Health care domain

Work Sponsorhip/Visa Required? : Yes, F-1 OPT


  • python
  • Machine Learning
  • Java


  • Teamwork
  • Networking
  • Management
  • Leadership
  • AWS
  • No SQL
  • Linux,MacOs,Windows
  • Git,SVN
  • iOS Development
  • Android Development
  • Tableau,Power BI


  • Data Scientist 02.00 Yrs
  • Machine Learning Engineer 02.00 Yrs
  • Data & Machine Learning Scientist 02.00 Yrs


Company Name: SCoRe Lab, University of South Florida

Job Title: Graduate Research Assistant

Experience: 2016 Aug - 2021 Aug (5yrs 1mo )

• Led a collaborative project and performed application development, implemented AI, troubleshoot systems, and handled users’ feedback. Currently, 45 patients with CHD history using our system and received grants amounting to $125,000

• Integrated machine learning model on the smartphone to detect user activity using sensor data with an accuracy of 87%

• Performed data analysis and applied computer vision techniques on image data to identify distracted driving classes with 94% accuracy

• Collected, processed, analyzed image data, and leveraged instance segmentation technique to identify mosquito breeding habitats

Company Name: BroTecs Technologies Ltd.

Job Title: Software Engineer(iOS)

Experience: 2014 Jun - 2016 Jul (2yrs 2mos )

• Developed multiple VOIP-based softphone application in the iOS environment for air to ground communication specifically for aviation domain

• Gathered user requirements, developed user interface, designed database schema, and modified voice protocol algorithm to enhanced voice call packet security

• Upgraded the application from objective-c to swift and integrated third-party/open-source library

• Collaborated with test team to conduct formative and summative evaluation and deliver the application to the client

Academic Projects

Title : Personalized Post-Cardiac Rehabilitation System
Technology Used : Java, Android, Sensor, Machine Learning, Sklearn

- Developed a smartphone and smartwatch application in the android environment to collect behavioral, physiological, and usage data

- Applied Random Forest model to detected real-time user activity (walking, running, sitting, etc.) using watch sensors data

- Integrated different google API services to upload data in the cloud and for instance message services

Title : Detecting Distracted Driving Activity from Image Data
Technology Used : Faster RCNN, Inception V2, ResNet-101, Python, NumPy, Pandas, Matplotlib, Seaborn

- Employed Faster RCNN technique to identify 9 types of objects in the image (e.g., phone, bottles, radio, hands, face positions) using Inception-V2 and ResNet-101 network

- Classify 10 types of distracted driving activities using different machine learning models and achieved 94% accuracy using the Random Forest model

Title : Detecting Distracted Driving using wrist-worn wearable sensor data
Technology Used : Python, Sklearn,, NumPy, Pandas, Matplotlib

- Collected sensor data using shimmer device and CAREN platform and extracted 14 features for the machine learning model

- Applied Naive Bayes, SVM, Decision Tree, and Random Forest to classify 4 classes and achieved 93.6% accuracy with RF model

Title : Identifying Mosquito Breeding Habitats
Technology Used : Mask RCNN, Python

- Collected video data using UAV, and preprocessed, analyzed data to identify breeding habitats

- Applied Mask RCNN technique to detect and localize the breeding habitats with 52.72 mean average precision


, Computer Science

, Computer Science

, Computer Science and Engineering