Every production based ML model has stories to tell that will not only engage, inform, surprise, delight, and impact their audience, but that will also deliver on measurable business goals. And I am the conduit between models and business goals.
I believe in continuous learning, standard work life balance, enhancing my skills with different programming language. Outside work, I love philosophy and psychology based topics. Someday I would love to work on former and latter integrated systems.
This repository contains the work related to deep-learning. Various algorithms, general framework of classification code, object detection and many other experiments.
In order to generate more images the latent space(bottleneck) consisting of vectors is altered. More can be found in the link
Data Science based experiment on LSTM Keras and visualisation techniques using Dash.
Small and periodically used python scripts for handy purposes.
Image Processing based script involving different functionalities like contours detection, blurring and rotating.
Computer Vision based motion tracking project that was implemented as a part of curriculum.
Simple automation saves a lot of time. This blog is explanation of my research project which involves application of ML algorithm to automate Matched Filter concept in Signal Processing. Click on the topic to get more insights about it.
This blog explains my Master thesis. The idea behind it is to use synthetic data to generate enough dataset to train a model. Later, the model is tested wih real-world data and performance is tested. Click on the topic to get more insights on this..
Separation of foreground and background images to produce different synthetic images. Subjective analysis on the image to check realism
What if we have less images and would like to generate more with slight changes in original ones. This blog tells more about it.
This part of the blog consists of explanation about variational autoencoder, math behind it, code and the result
The method used to perform the Kaggle competition is explained briefly. Click on the topic to get more insights about it.
Out of many sleepless errors I have encountered, the CuDA issue to run the deep learning training really had caused big problem. A small blog on how it can be handled easily. Click on the topic to get more insights about it.
Series of basics of machine learning - part 1
Series of basics of machine learning - part 2
Series of basics of machine learning - part 3
Series of basics of machine learning - part 4