AI-POWERED VIDEO VITALS DETECTION SYSTEM

I developed an advanced AI-Powered Vitals Detection System that analyzes a 20-second facial video to estimate essential health vitals with high accuracy. The system leverages computer vision, deep learning, and signal processing to extract real-time physiological and biometric insights from facial cues.

🔹 Key Features:

✔️ Detects multiple vitals such as Heart Rate, Respiratory Rate, SpO₂, Stress, Age, Gender, BMI, Jaundice indicators, Facial Symmetry, Pupil Dilation, and more.
✔️ Uses state-of-the-art deep learning models for age, gender, BMI, and stress prediction.
✔️ Utilizes rPPG and Eulerian Video Magnification techniques to analyze subtle skin tone variations for heart rate and SpO₂ estimation.
✔️ Extracts and processes video frames to compute accurate health parameters within seconds.
✔️ Designed for high accuracy, robustness, and real-world usability.
✔️ Includes a clean, user-friendly UI interface for smooth interaction and report generation.

This project highlights my expertise in computer vision, deep learning, biomedical signal processing, and AI-driven health analytics — bringing cutting-edge research into practical, real-world application.

View Documentation

DEEP FAKE DETECTION SYSTEM

I developed an advanced Deepfake Detection System using a hybrid deep learning model to identify manipulated videos with high accuracy. The system is built on top of CNN and GRU layers, leveraging powerful architectures like DenseNet121, EfficientNet-B1, InceptionV3, and Xception.

🔹 Key Features:

✔️ Utilizes state-of-the-art deep learning models for enhanced detection.
✔️ Integrates CNN for feature extraction and GRU for temporal analysis of video frames.
✔️ Processes videos frame-by-frame to detect manipulated content effectively.
✔️ Designed for high accuracy and efficiency in deepfake identification.
✔️ Comes with a user-friendly UI interface for seamless interaction.

This project showcases my expertise in computer vision, deep learning, and AI-based video analysis while tackling the critical challenge of detecting synthetic media.

Read Research Paper

FACIAL SPOOFING DETECTION USING CNN

🔹 Key Features:

✅ Advanced CNN Architectures:
✔️ Implemented and compared AlexNet, GoogleNet, and ResNet for facial spoof detection.
✔️ Achieved the highest accuracy with AlexNet, demonstrating its efficiency in detecting presentation attacks.

✅ Robust Dataset & Training:
✔️ Created a dataset with real and spoofed face images using diverse attack scenarios.
✔️ Applied data preprocessing, augmentation, and noise injection to enhance model generalization.
✔️ Split the dataset into training, validation, and testing sets for optimal performance evaluation.

✅ Liveness Detection & Real-Time Testing:
✔️ Designed a model to accurately classify live vs. spoofed faces.
✔️ Evaluated the model with a webcam-based real-time liveness detection system.

✅ Performance Evaluation:
✔️ Used metrics like Precision, Recall, F1-score, and AUC-ROC to assess model effectiveness.
✔️ Conducted a comparative study, identifying the best-performing CNN model.

✅ Technology Stack:
✔️ Deep Learning Frameworks: TensorFlow / PyTorch
✔️ Programming Language: Python
✔️ Models Used: AlexNet, GoogleNet, ResNet

This project significantly enhances biometric authentication security, reducing vulnerabilities in facial recognition systems by effectively detecting spoofing attempts.

See Research Paper

PHARMACY MANAGEMENT SYSTEM

🔹 Key Features:

✅ Inventory Management:
✔️ Tracks medicine stock levels, expiration dates, and batch details.
✔️ Alerts for low stock and expired medicines to prevent shortages and wastage.

✅ Sales & Billing System:
✔️ Generates automated invoices and bills for seamless transactions.
✔️ Supports customer purchase history tracking for better service.

✅ Prescription & Customer Management:
✔️ Maintains digital records of prescriptions and patient details.
✔️ Provides a searchable database for quick retrieval of past prescriptions.

✅ Database & Security:
✔️ Designed with efficient relational database models (ER diagrams, SQL queries).
✔️ Implements data validation and security measures to prevent unauthorized access.

✅ Technology Stack:
✔️ Database: MySQL / PostgreSQL
✔️ Backend: PHP / Python (Django, Flask)
✔️ Tools: XAMPP, phpMyAdmint

This system ensures efficient pharmacy operations, reduces manual errors, and enhances customer service by automating critical processes.

Linkedin

If you'd like to get in touch, please feel free to connect with me on LinkedIn.

Instagram

Feel free to connect with me on LinkedIn.

Medium

I'm a Data Scientist and AI Engineer enthusiast who enjoys writing on Medium. I share my experience, insights and lessons to help others learn. As I continue to work in this field, I aim to help build a global community where enthusiasts join forces and engage in meaningful conversations.