Traffic congestion Prediction - Time Series

Predicted the future traffic congestion in different junction of road in real time data Performed EDA to analyse the data Tested models like ARIMA, SARIMA, Random Forrest, LightGBM and LSTM, GRU Forecasted the congestion on data by Random Forrest with 76.6% accuracy on avg across junctions

Background Generation - Generative AI

Created pipeline of object detection, segmentation and then image generation using Stable Diffusion inpainting. Performed Object detection on the basis of prompt with the help of Grounding Dino model Generated bounding box coordinates are used to create mask with the help of Segment Anything model

Sentiment Analysis and Fine Tuning with LoRA

Here in this project I have explore the Fine Tuning with LoRA and tried to perform Sentiment Analysis on my data. Sentiment analysis is a NLP technique used to determine the sentiment or opinion expressed in a piece of text. It categorizes the emotions conveyed is positive or negative.