Nikhil Thakurdesai

Papers

Total Citations

H-Index

About

Nikhil Thakurdesai is a researcher at the forefront of computer vision, specializing in semantic and few-shot segmentation for challenging visual domains. His most-cited work, "Few-Shot Segmentation and Semantic Segmentation for Underwater Imagery" (2023, 7 citations), addresses a critical bottleneck in marine AI: the scarcity of diverse, annotated underwater datasets. To overcome this, Thakurdesai introduced a novel, animal-centric dataset with dense pixel-level annotations spanning fine-grained marine categories—a foundational resource that enables models to learn from minimal examples. This contribution directly advances autonomous underwater exploration, ecological monitoring, and marine conservation by making segmentation models more adaptable to rare or unseen species. Beyond dataset creation, his work demonstrates robust segmentation pipelines that generalize across domains, bridging the gap between few-shot learning and real-world deployment. By tackling both data scarcity and model generalization, Thakurdesai’s research empowers AI systems to interpret complex underwater scenes with unprecedented precision, laying critical groundwork for scalable, data-efficient vision in extreme environments. His efforts are shaping the next generation of intelligent marine robotics and environmental sensing.