The objective is to build various classification models, tune them and find the best one that will help identify failures so that the generator could be repaired before failing/breaking and the overall maintenance cost of the generators can be brought down.

Up-sampling and denoising signals using a deep neural network model

Developed a novel upsampling algorithm which accurately identifies impactful rows from the dataset

a complete classification analysis based on multiple models on census data

Data Mining and Machine Learning APS Failure at Scania Trucks Data Set.

Upsampling already available land cover raster layers using machine learning inside Google Earth Engine platform.

A Novel Methodology of Domain Wise feature selection approach which is capable of identifying the interrelationships by focusing on Domain-Wise feature selection. It ensures that correlated and similar features are considered together by grouping them in similar domains based on correlation values

An implementation of SMOTE

This project aims to analyze diabetes data using data management, captivating visualizations, and cutting-edge machine learning techniques to predict the presence of diabetes in individuals. Our robust dataset includes comprehensive health exam results and family history.

3D Sensing and Sensor Fusion

This is an implementation of the CVPR 2017 paper - Deep Koalarization : Image Colorization using deep CNNs and Inception Resnet V2

This work predicts the buyers and non-buyers on an online shopping platform at a 92.4% accuracy, 89.8% precision and 95.3% recall performance.

Dive into the world of Signal and Image Processing with this repository. Explore a collection of Python programs covering Discrete Fourier Transform, Elementary Signals, Sampling, Point Processing Techniques, Histogram Processing, Frequency Domain Filtering, Edge Detection, Erosion and Dilation, and Morphological Operations.

Using ML models to classify fraud transactions

Handy DSP routines

IEEE TCSVT_2024_PU-Mask: 3D Point Cloud Upsampling via an Implicit Virtual Mask

Oversample 44.1kHz WAV file to 352.8kHz

“PUFA-GAN: a frequency-aware generative adversarial network for 3D point cloud upsampling,” IEEE Transactions on Image Processing, vol. 31, pp. 7389 – 7402, Nov. 2022.

MNIST Image reconstruction using Autoencoders