Soumyadipta Kar

Location: Kolkata, West Bengal, India
Phone: +91-XXX0871426
Email: soumyadipta.kar2002@gmail.com
LinkedIn: soumyadipta-kar2002
GitHub: SoumyadiptaKar

---

Education

Haldia Institute of Technology — Haldia, West Bengal, India

B.Tech. in Computer Science and Engineering | 2020 - 2024

• CGPA: 9.19

Calcutta Public School — Kolkata, West Bengal, India

Higher Secondary Education (ISC) | 2020

• Percentage: 89.75%

Calcutta Public School — Kolkata, West Bengal, India

Secondary Education (ICSE) | 2018

• Percentage: 88%

---

Skills & Coursework

• Data Structures & Algorithms
• Operating Systems
• Database Management Systems (DBMS)
• Object-Oriented Programming (OOPS)
• Artificial Intelligence
• Machine Learning
• Deep Learning
• Computer Vision
• Natural Language Processing

---

Internships & Training

Indian Institute Of Technology Bhubaneswar

Junior Research Intern | Jun-Aug 2023 | Bhubaneswar, India

• Collaborated with the data team to pre-process video datasets for classification tasks, resulting in a 15% increase in model accuracy.
• Designed, developed, and optimized classification models utilizing cutting-edge video transformers, C3D, I3D, and other deep-learning architectures.
• Implemented efficient training pipelines to train video classification models, reducing training time by 20%.

National Institute Of Technology Durgapur

Summer Intern | Jul-Aug 2022 | Durgapur, India

• Designed and implemented a robust methodology for selecting the most suitable ARIMA framework, ensuring accurate predictions of future income for consecutive years.
• Maintained detailed documentation of experiments, methodologies, and results.

Indian Statistical Institute

Student Trainee | Jan-Mar 2022 | Kolkata, India

• Engaged in hands-on learning of deep learning principles, from the fundamentals of perceptron and neural networks.
• Demonstrated understanding of forward and backward propagation, activation functions, and optimization techniques through practical exercises.
• Participated in seminars, workshops, and networking sessions with esteemed faculty, researchers, and industry experts.

Indian Institute of Technology Kharagpur

Student Trainee | Sep-Dec 2023 | Kharagpur, India

• Utilized mathematics, Python programming, and intuition in machine learning problem-solving to develop a comprehensive mental model for data science challenges.
• Applied theoretical knowledge in real-world scenarios through hands-on projects in machine learning.
• Awarded Certificate of Merit for securing a position in the top 2%.

---

Academic Projects

Determination of Heat Flux using 3D Convolution

Technologies: I3D, C3D | Jun-Aug 2023

• Developed a novel method to accurately estimate heat flux patterns from video datasets using 3D convolutional neural networks (CNNs).

Chicken Disease Classification

Technologies: VGG-16, AWS | Sep-Nov 2023

• Created a tool for early detection of coccidiosis in chickens using image analysis, aiding in effective disease management and prevention.

Text Summarization

Technologies: Transformers, AWS | Dec 2023-Feb 2024

• Implemented dialogue summarization using Google’s Pegasus model, leveraging the Samsum dataset for summarization tasks.

Exoplanet Detection

Technologies: SMOTE, ADASYN | Dec 2023-May 2024

• Utilized deep learning models to identify exoplanets based on the periodic dimming of stars caused by planetary transits, using data from the Kepler Mission.

---

Research Publications & Presentations

Automated Prediction of Income from Farming of a Commodity: An ARIMA based framework

Technologies: ARIMA

• Proposed an efficient ARIMA-based framework for predicting farmers’ income in future years.
• Presented at the International Conference on Optimization, Learning, and Analytics Business (Dec 2022).
• Published in the Journal of Decision Analytics and Intelligent Computing on 21 July 2023.
• Publication

Efficient Healthcare Supply Chain: A Prioritized Multi-task Learning Approach with Task-specific Regularization

Technologies: NN, MTL

• Improved healthcare supply chain operations through a robust machine learning approach.
• Objectives included predicting “shipping quantity” and “time to delivery” and identifying key factors influencing healthcare supply management.
• Published in Engineering Applications of Artificial Intelligence on 14 March 2024.
• Publication

DeepCrop: An Efficient Deep Learning Approach for Monitoring Crop Growth

Technologies: Transformers

• Developed an automated tool for monitoring crop growth using Transformer-based models to capture temporal patterns.
• Presented at the 3rd International Conference on Nonlinear Applied Analysis and Optimization (ICNAAO-2023).

AI-Driven Methodology for Prediction of Nucleate Pool Boiling Heat Transfer over Vertical Tubes (Communicated)

Technologies: I3D, C3D

• Predicted nucleate boiling heat transfer on vertical tubes using machine learning and high-speed video data.
• Developed an I3D CNN model for estimating heat flux and conducted a comparative analysis with C3D and ResNet3D models.
• The model achieved a mean error of 6% in heat flux prediction.

---

Technical Skills

• Languages: Python, Java, C, SQL
• Developer Tools: VS Code, Jupyter Notebook, Google Colab, IntelliJ IDEA Ultimate
• Technologies/Frameworks: Linux, GitHub, TensorFlow, PyTorch, Transformers, Docker, AWS