Pramod Kompalli

Background

About Pramod Kompalli

I am interested in the broad area of Machine Learning, especially when applied to Computer Vision and Multimedia Understanding problems. I'm currently exploring Deep Learning solution stacks to address some previously unsolvable problems in multi-modal data fusion. I'm particularly keen on understanding why Deep Learning solutions work and what their limits are. Previously, during my PhD, I proposed new approaches to scalable retrieval/recognition. I built retrieval systems over movies/TV shows, Cricket videos, and large document image databases. I built the biggest retrieval system over non-English document images, whose performance defined the state-of-the-art at the time. My Thesis won the Doctoral Dissertation Award by IUPRAI. I regularly review for several computer vision and document recognition conferences and journals. I have received the outstanding reviewer award three times. I enjoy working on problems of scale and real-world impact. I hope to be involved in building and deploying a billion-scale visual understanding system. I wish to learn how to build products, systems and have a positive impact on customers.