KRISHNA VAMSI ROKKAM's Projects
Most people are currently suffering from various skin allergies after recovering from COVID-19. Recent researchers with the help of dermatologists stated that there is an association between skin problems and COVID-19. It has become a daunting task to categorize the skin changes visually without any help of a skin specialist and also treatment procedures for these kind of diseases are obscure. Therefore, our problem statement revolves around the idea of classifying the skin disease severity so that the spread of disease can be identified in the early stages and the worsening of disease can be prohibited.
We developed an AI/ML-based solution that generates short summaries of press releases for easy and more targeted dissemination of information to the people such a TL, DR (Too Long, Didn’t Read)
This repository hosts the complete Text-to-Speech (TTS) system developed as part of the COMP691 course, showcasing the practical application of advanced machine learning techniques to generate human-like speech.
The Leek group guide to data sharing
Conversational AI, RNNs, Transformers, Speech Brain, Speech Recognition, Machine Translation, Sequence to Sequence Learning, Self supervised learning, Autoencoders, GPT, Fine Tuning LLMs
Performed Predictive Modeling on the wine dataset and built five models for comparing the accuracy of predicting the quality of wine into categorical values. the highest accuracy was obtained using the XG Boost algorithm among Random Forest, Decision Tree, Gradient Boost, and Ada Boost models
This project deals with handwritten color digits by performing two tasks at a time i.e by predicting the value of the digit as well as the color of the digit. I created a data generator function that generates red, green color images using the greyscale MNIST images dataset from Keras. Here, I used Resnet style architecture by using skip connection with Add Layer in order to create a multi-tasking CNN model. The generated color images are fed to the model that is done in multi-tasking mode by implementing two CNNs for digit and colour recognition at an accuracy of 98%.
OpenTracks is a sport tracking application that completely respects your privacy.
Stock market prediction is an attempt of determining the future value of a stock traded on a stock exchange. This project focuses on classification problems, predicting the next-second price movement, and acting upon the insights generated from our models. We implemented multiple machine learning algorithms including logistic regression, support vector machines (SVM), Long- Short Term Memory (LSTM), and Convolutional Neural Networks (CNN) to determine the trading action in the next minute. Using the predicted results from our models to generate the portfolio value over time, the support vector machine with a polynomial kernel performs the best among all of our models.
The model has been developed with good accuracy using CNN, to create a realistic and user-friendly model to easily track the Body Mass Index of users which helps determine the fitness and assess the body fat by nutritionists and have a fit and healthier life.
The project deals with Detecting skin diseases based on images. The model has been implemented using Python and Convolutional Neural Networks and OpenCV. The approach works on color images and greyscale images. Used different Neural Network layers such as Max-Pooling, Flatten, Conv2D, etc. to build a system that successfully detects skin diseases based on images captured through camera and deployed model using flask application and web development technologies. Received Silver Award at Ennovate-The International Innovation Show-2021, Poland for this innovation
APP Project
A repository that will link with RStudio
This project deals with the concepts of Natural Language Processing, where the raw data is transformed into useful text data by lowering the case, removing stopwords, and applying stemmer algorithm to the text data, and finally applied Vectorization techniques. The model has given better accuracy using the Support Vector Classifier algorithm on the processed input text data by converting the text into spam and non-spam categories.