Hello, I am Priyanka Moorthy.

GAN network for image translation from line art to colored butterfly images with a Pix2Pix Architecture. Achieved discriminator and generator loss of 0.91 and 1.23 respectively.

Image segmentation model built using Mask-RCNN architecture on nuclei data to perform object detection and segementation masks on nuclei in images. Achieved a mean average precision of 0.73.

Working on implementing a modified SAM-SLR framework for continous sign language recognition in realtime as part of my Master's project

Scraped thousands of movie records including plot line and storyline features from the IMDB website using BeautifulSoup. Performed data cleaning, removed stop words, applied PorterStemmer to stem words. Fine-tuned a Pegasus model that follows a Self-Supervised Objective.

Data preprocessing, word embedding, and experimental analysis of sentiment classification on amazon reviews data by implementing FNN, CNN, LSTM, and distillBERT models.

Performed data cleaning, Exploratory Data Analysis and mined answers from data for proposed research questions

Data Collection, Visualize features extracted by ResNet101 network using t-SNE

Implemented a sequential Deep-Q Learning on Open AI’s continuous mountain car environment that converges in 1643 episodes with 12 hidden dimensions.

Trained a LSTM model on Lo-Fi style music data to generate similar music.