George Mason University [Aug 2020-till date]
- Built a deep learning framework that combines different color spaces trained on multiple CNNs and gives a global classification using fusion strategies for liveness detection. Our framework consists of Alexnet, Vgg(16,19), GoogLeNet, ResNet(18,34,50), DenseNet-121 and MobileNet V3.
- As part of our research, I built a personalized U-Net and Faster-RCNN deep learning segmentation framework for Contactless fingerprint processing.
- Intern at Accure Analytics: Developed a Handwriting image segmentation algorithm using various Image processing techniques like the moving window algorithm and the Otsu automatic global threshold detection algorithm, annotated the segmented images, and the annotations are used for training the LSTM model for Optical character recognition (OCR).
- Built multiclass classification machine learning models using KNN, MLP, and RFC on Patient Discharge Location using MIMIC-IV | Key Contributions: data handling, Integration, modeling, visualizations, feature importance.
- I worked on a project where I used various data analysis techniques to find how features like state holidays, temperature, and precipitation affect beer consumption in Sao Paulo.
George Mason University [Aug 2019-Jan 2020]
- I am working on PAD using various deep learning and image processing techniques to classify live and spoof finger images in different color space domains.
- Worked as Graduate Research Assistant under Dr Michael Von Fricken, we modeled a stationary robot that controls lawn ticks, I built and studied the robot performance and was collecting data from various sensors in the outside environment.
- Worked on data handling and modeling data visualizations and projecting better performance graphical representation on multiple datasets
- Working on machine learning and Deep learning Algorithms like support Vector Machine and Neural Networks.
- Worked on small and medium datasets and modeled them using Perceptron linear-gradient descent algorithms.
JNTUH College of Engineering [Jan-2019- May 2019]
- Worked on the hybrid model by simulating Solar and Wind energy and their Harmonic reduction by connecting the mini-grid from domestic houses to the main grid using Matlab and Simulink.
- Worked on harmonic reduction using D-Statcom and also successfully simulated the model using Simulink Matlab.
Defense Research & Development Laboratory, India.
- Theoretical, Experimental Comparison and Mathematical Modelling of AC-AC 3-Phase Thyristor.
- Detailed Study over 10 different Electrical Transducers in the Structural Test Facility DRDL.
- Project on Open-Loop Heat Flux Control on Metallic Sample Plate and observation on how missile Fins are tested using this process.