Hi 👋, I’m Vighnesh Nayak

A passionate Mechanical Engineering student exploring the intersection of ML, Robotics, and IoT

👨‍🎓 Education

• B.Tech in Mechanical Engineering, Indian Institute of Technology Bombay (2021-2025)
• Minor: Artificial Intelligence and Data Science

🏆 Achievements

• Among top 2.3 percentile in IIT JEE-Advanced
• Ranked 123 in K-CET (2021)
• Second place in Astromania-2022 quiz by Krittika - Astronomy club
• Mentored students in Computer Vision at IIT Bombay

💼 Professional Interests

AI | Machine Learning | Robotics

🚀 Featured Projects

Sign Language Video to Audio Conversion

• 🎥 Developed a system to convert sign language videos into audio outputs without relying on text.
• 🔍 Implementation Details:
  • Utilized a modified I3D model for feature extraction from video data.
  • Processed the WLASL-2000 dataset to generate spectrograms using Tacotron 2 and Hifi GAN models.
  • Used Non-Maximal Suppression for continuous sign identification.
• 📊 Results:
  • Achieved effective translation of sign language into corresponding audio descriptions.
  • Enhanced model accuracy through extensive data preprocessing and augmentation.
• 🛠️ Tech Stack: Python, Pytorch, PytorchVideo, Librosa, Numpy
• 🔗 Project Repository

Musical Instruments Separation Using Deep Neural Networks

• 🎵 Implemented audio source separation using deep learning techniques
• 🔍 Implementation Details:
  • Utilized MUSDB-18 dataset with 44100Hz to 8192Hz downsampling
  • Performed Short Term Fourier Transform (STFT) for spectrogram generation
  • Implemented U-Net architecture for semantic segmentation
  • Achieved high-quality separation of instruments and vocals
• 🛠️ Tech Stack: Python, TensorFlow, Numpy, Librosa
• 📈 Results: Successfully separated multiple instrument tracks with minimal artifacts
• 🔗 Project Repository

Local Moodle: Automated Moodle Scraper

• 📥 Developed a Python script to automate the downloading of files and posts from Moodle courses.
• 🔍 Implementation Details:
  • Utilizes Selenium and WebDriver for browser automation.
  • Scrapes course information from the Moodle homepage, creating organized folders for each course.
  • Downloads course materials (e.g., PDFs) that haven’t been previously downloaded.
  • Extracts and saves forum posts along with any attachments.
• 📊 Results:
  • Streamlined the process of accessing and organizing course materials.
  • Enhanced user experience by automating repetitive tasks associated with Moodle.
  • Tailored specifically for use with the Moodle version at the Indian Institute of Technology Bombay (IITB).
• 🛠️ Tech Stack: Python, Selenium, BeautifulSoup
• 🔗 Project Repository

Joint Dictionary Learning for Color Image Demosaicing

• 📸 Developed an advanced image processing algorithm without prior training data
• 🔍 Implementation Details:
  • Used Gradient Corrected Bilinear Interpolation (GCBI) for initial estimates
  • Developed patch-based processing pipeline
  • Implemented joint sparse demosaicing dictionary learning
• 📊 Results: Achieved 6.67% mean log PSNR improvement over GCBI
• 🛠️ Tech Stack: Python, OpenCV, Numpy
• 🔗 Project Repository

Cargo Bots: Swarm Robotics System

• 🤖 Built a warehouse simulation model using coordinated micro-controller bots
• 🔍 Implementation Details:
  • ArUco marker-based localization using overhead camera
  • A* algorithm for priority-based path planning
  • Collision-free trajectory generation
  • PID feedback control implementation on ESP32
• 📈 Results: Successfully demonstrated autonomous cargo transport with multiple bots
• 🛠️ Tech Stack: Python, OpenCV, ESP32
• 🔗 Project Repository

IoT Monitoring System (Zwilling Labs Internship)

• 📊 Developed enterprise-grade IoT monitoring solution
• 🎯 Key Features:
  • Real-time event triggering system
  • Optimized schema design for time-series data
  • SQLite-based event tracking
  • Interactive data visualization
• 🛠️ Tech Stack: Python, PostgreSQL, TimescaleDB, Svelte, SQLite
• 📈 Impact: Improved data retrieval performance by 40%

Topology Optimization for Robotic Gripper

• 🦾 Developed optimization framework for robotic gripper design
• 🔍 Implementation Details:
  • FEniCS-based topology optimization
  • SIMP algorithm for material distribution
  • 3D visualization using PyVista
  • Stress and strain analysis
• 📊 Results:
  • 50% reduction in material usage
  • Maintained structural integrity
  • Optimized force distribution
• 🛠️ Tech Stack: Python, FEniCS, PyVista
• 🔗 Project Repository

🛠 Skills

Languages

Python | C++ | JavaScript | SQL

Libraries

Numpy | Pandas | Scikit-Learn | Keras | TensorFlow | OpenCV

Tools

Git | LaTeX

Operating Systems

Windows | Linux

📊 GitHub Stats

🤝 Let’s Connect!