*

here are some of the publications, patents, and research projects i've contributed to, driven by a passion for solving real-world problems.
I like exploring and conducting research in areas of AI ethics.

Cuffless BP Estimation

Deep Learning, Machine Learning, SWIN Transformers | Published

Developed cuffless blood pressure estimation system using bispectrum and bicoherence images for accurate predictions, with a SWIN Transformer model for non-invasive monitoring.

achievements

  • Implemented bispectrum and bicoherence image generation for physiological signal processing
  • Trained a SWIN Transformer for robust, non-invasive BP estimation
  • Published work demonstrating high prediction accuracy for clinical applications

technologies

pythondeep learningmachine learningsignal processingswin transformer

A System and Method for Automated Waste Segregation

Patent | Application no: 202441064693

An AI-powered machine that automates construction waste segregation, enhancing recycling efficiency and sustainability by diverting waste from landfills and integrating with waste management systems.

achievements

  • Designed intelligent vision system for construction waste classification
  • Integrated real-time sorting mechanisms to improve recycling efficiency
  • Patented technology for automated waste segregation and landfill diversion

technologies

computer visionmachine learningembedded systemsrobotics

Federated Learning for Privacy-Preserving AI

Research in Progress (Ongoing)

Ongoing research on privacy-preserving machine learning models using Federated Learning to enable collaborative training without sharing raw data.

achievements

  • Developing a federated learning pipeline for distributed environments
  • Integrating differential privacy to protect sensitive user data
  • Exploring cross-device model aggregation and optimization techniques

technologies

pythonfederated learningdifferential privacypysyftpytorch