Lalith Sharan

Hi, welcome to my website! I'm Lalith, a researcher in the field of computer vision, with a focus on computer assisted surgeries. My PhD centres on deep learning based endoscopic image analysis. Here, I worked on unpaired image translation, landmark detection, 3D reconstruction, depth estimation, and segmentation.

My work contributed to 15+ publications in high-impact journals (IEEE JBHI, IJCARS, Annals of Thoracic Surgery, EJCTS, pre-work for Nature) and conferences (MICCAI, SPIE, CURAC, BVM), garnering 6 poster awards and invited talks.

Besides training models, I love science communication - I have presented my research work as science slams in locations across Germany for a cumulative audience of 5000+, earning two nominations to the German Science Slam Championships 2023, 2024. In my free time, I love to nerd out on science history, mathematics, cook, or write. Check out my research and publications, or just get in touch with me for my resume or to know more!

News

Oct 2024

I have successfully defended my doctoral degree (Dr. sc. hum.) on the 16th of October, 2024! The PhD was graded with a summa cum laude, pending external review.

Aug 2024

My cumulative dissertation titled Deep learning based image analysis for endoscopic minimally invasive mitral valve repair is successfully accepted by the doctoral committee! The examiners for my defense have been assigned.

Jun 2024

I gave a hands-on tutorial on transfomer models for medical computer science students.

Jun 2024

I gave two science slams: in Heidelberg (first place) and Mainz (shared first place)