Linbo Qing

Papers

Total Citations

H-Index

About

Linbo Qing is a researcher whose work lies at the intersection of computer vision, social intelligence, and efficient deep learning. His key research areas include social relation recognition and neural network compression, where he has made notable contributions to both understanding human interactions and optimizing AI models. In his highly cited 2022 work, "Multi-Level Transformer-Based Social Relation Recognition," Qing introduced a novel transformer architecture that captures hierarchical social cues, enabling more nuanced recognition of interpersonal dynamics—a critical step toward building sophisticated social intelligent systems. This paper has already garnered 11 citations, reflecting its growing influence in the field. Additionally, his 2020 paper, "Re-training and parameter sharing with the Hash trick for compressing convolutional neural networks," addresses the practical challenge of deploying deep models on resource-constrained devices, achieving efficient compression without sacrificing accuracy. By combining theoretical insight with algorithmic innovation, Qing’s research bridges the gap between high-level social understanding and low-level computational efficiency, offering valuable tools for both academic study and real-world applications. His work continues to inspire advances in socially aware AI and model optimization.