Machine Learning Computer Vision Intern Candidate

Soubhik Majumdar

San Jose, California, USA US Citizen

********************@******.*****.*** com/ SoubhikMajumdar linkedin.com/in/soubhik-sinha-majumdar Education

San Jose State University San Jose, CA

Master of Science in Electrical Engineering Aug 2024 – Present GPA: 3.89/4.0

• Specialization: Artificial Intelligence, Machine Learning, and Deep Learning PES University Bangalore, India

Bachelor of Technology in Electronics and Communication Engineering Aug 2018 – May 2022 GPA: 3.78/4.0

• Specialization: Signals and Systems

Work Experience

Nybsys Inc San Jose, CA

Software Development Intern Jul 2025 – Aug 2025

• Built a real-time CRM platform using FastAPI, Next.js, and WebSockets with live synchronization across multiple clients

• Designed and implemented scalable REST APIs and data pipelines for Contacts, Leads, Opportunities, and Deals

• Modeled relational data using PostgreSQL and SQLAlchemy ORM with optimized joins and indexing

• Improved system performance and concurrency handling through query optimization and batched real- time updates

Projects

3D LiDAR Birds’-Eye-View Object Detection in Urban Driving Scenarios Sep 2025 – Dec 2025 San Jose State University

• Developed 3D LiDAR Point Cloud based Bird’s-Eye View (BEV) Object Detection framework using fine-tuned BEVFusion

• Implemented inference, visualization, fine-tuning, and evaluation pipelines on nuScenes dataset with 60.5% mean Average Precision

• Technologies: LiDAR, Computer Vision, Object Detection, Point Cloud Library (PCL), Perception Systems

Deep Reinforcement Learning for 6G RIS Optimization Aug 2025 – Present San Jose State University

• Developing offline Deep Q-Network (DQN) for discrete RIS phase optimization in 6G MIMO systems

• Implemented physics-based dataset learning to optimize RIS configurations without environment inter- action

• Achieved optimal discrete actions to maximize SINR and user fairness in wireless systems

• Technologies: Deep Reinforcement Learning, 6G Wireless Communications, DQN, MIMO Systems Skin Lesion Classification DCNN for Melanoma Detection Mar 2025 – May 2025 San Jose State University

• Built deep convolutional neural networks for multi-class skin lesion classification using ISIC2018 dataset

• Achieved 67% accuracy across all lesion classes with focus on Melanoma detection precision

• Improved recall and precision for skewed minority classes through data augmentation, hyperparameter tuning, regularization, and architecture optimization

• Technologies: Deep CNNs, Supervised Learning, Model Training, Medical Image Analysis Swimmer Segmentation and Pose Estimation using SAM and YOLOv7 Aug 2024 – Dec 2024 San Jose State University

• Developed framework for segmentation and pose estimation on 4 video datasets of underwater swimmers from SJSU swimming team

• Utilized Meta’s Segment Anything Model (SAM) and YOLOv7 object detection for accurate joint tracking

• Designed and implemented script to identify, track, and display key joints (elbows, hips, knees) with high accuracy

• Technologies: Image Segmentation, Object Detection, YOLOv7, SAM, Pose Estimation Optimal Beamformer Design for mm-Wave Systems Jan 2021 – Sep 2021 PES University

• Developed fully connected multi-layer feed-forward deep neural network to optimize analog phase shift values in mm-Wave communication system

• Trained and tested Beamforming Neural Network model to maximize sum-rate in 2-user and 3-user multiple antenna systems

• Obtained sum-spectral efficiency vs SNR characteristics under different CSI errors

• Technologies: Neural Networks, Wireless Communications, mm-Wave Systems, Beamforming Gesture Controlled Rover Using Image Processing Jan 2020 – May 2020 PES University

• Implemented real-time gesture recognition system using Hough Transform and Contour Detection

• Extracted positional and motion data from mounted camera to control rover orientation

• Translated gesture inputs into control signals for camera angle adjustment and DC servo motor actua- tion

• Technologies: Image Processing, Computer Vision, Real-time Systems, Embedded Systems Technical Skills

ML & AI: Probabilistic Models & Statistics, Neural Networks, Deep Learning, Computer Vision, Rein- forcement Learning, Model Optimization & Evaluation Vision: Image Preprocessing, Image Segmentation, Object Detection & Classification, 3D Point Cloud Processing, Vision Transformers, LiDAR, Bird’s Eye View Perception Languages: Python, C++, MATLAB, SQL

Frameworks: PyTorch, TensorFlow, Keras, scikit-learn, OpenCV, YOLO, CUDA, FastAPI, Next.js Libraries: pandas, NumPy, SciPy, Point Cloud Library (PCL), SQLAlchemy, WebSockets Tools: Git, Google Colab, WSL2, PostgreSQL

Contact this candidate