Tali Treibitz

Papers

Total Citations

H-Index

About

Tali Treibitz is a pioneering researcher at the intersection of computer vision, marine science, and underwater robotics. Her work addresses some of the most technically demanding challenges in visual perception — specifically, how machine learning and imaging systems can be adapted to function in the complex, light-distorting underwater environment. Among her most recognized contributions is "CoralSeg" (2019), which has accumulated 71 citations and introduced an innovative approach to semantic segmentation of coral reef imagery using sparse annotations. This work is particularly significant because it dramatically reduces the laborious manual labeling typically required for biological survey data, enabling more scalable and automated analysis of fragile marine ecosystems. Her more recent work on self-supervised monocular depth estimation underwater (2023) tackles the fundamental robotics challenge of spatial awareness beneath the ocean's surface, where conventional depth estimation methods fail due to light attenuation and color distortion inherent to aquatic media. Treibitz's research sits at a vital crossroads: advancing core computer vision methodology while directly serving environmental monitoring and conservation applications. Her contributions empower marine biologists and autonomous underwater vehicles alike, making sophisticated AI tools accessible in one of Earth's least understood and most ecologically critical environments.