
The Segment Anything Model (SAM) can effectively segment the pupil and iris in eye images captured in virtual reality setups, but struggles to accurately segment the sclera without extensive manual guidance.


coremsg

segment-anything-model-demonstrates-promising-zero-shot-segmentation-of-eye-features-in-virtual-reality-setups


Segment Anything Model Demonstrates Promising Zero-Shot Segmentation of Eye Features in Virtual Reality Setups
