Summer Research Intern
Preserving Topology Using Deep Image Segmentation under the supervision of Dr Amit Chattopadhyay. Introducing a continuous-valued loss function to preserve correct Topology of fine-scaled structures. This function is designed to enforce segmentation to maintain correct topology akin to the ground truth, ensuring consistent Betti numbers. The innovative topology-preserving loss function is differentiable, allowing seamless integration into end-to-end training of deep neural networks. Four different methods are used to preserve the Topology of CREMI-A dataset. I am also profoundly grateful to the IASc-INSA-NASI Summer Research Fellowship Program for theirgenerous support and for providing me with the opportunity to pursue this research