Krishna Agarwal

The paper proposes finetuning the Segment Anything Model (SAM) using large-scale synthetic fluorescence microscopy data to achieve robust and high-performing instance segmentation of mitochondria, addressing the lack of real annotated datasets.