Yulong Liu

Papers

Total Citations

H-Index

About

Yulong Liu is a researcher whose work lies at the intersection of computer vision and unsupervised feature learning, with a particular focus on multimodal data integration. His most-cited paper, "Unsupervised multimodal feature learning for semantic image segmentation" (2013, 33 citations), introduces a novel approach that leverages single-layer networks to learn features from both RGB and depth images without requiring labeled data. A key innovation of this work is the strategic selection of dictionaries from existing data, enabling more effective and efficient feature extraction for semantic segmentation tasks. This contribution addresses a fundamental challenge in computer vision: how to combine complementary visual modalities in an unsupervised manner. By demonstrating that meaningful representations can be learned from raw multimodal input, Liu's research has implications for autonomous systems, robotics, and scene understanding where labeled data is scarce. His work stands out for its elegant simplicity—using shallow architectures to achieve robust performance—and has influenced subsequent research in unsupervised multimodal learning. Liu's approach continues to be relevant for researchers seeking data-efficient methods for semantic segmentation in real-world applications.