John Peruzzi

Papers

Total Citations

H-Index

About

John Peruzzi is a researcher at the intersection of computer vision and social signal processing, with a focus on understanding human social interactions from visual data. His primary research areas include conversational group detection, social gathering analysis, and deep learning for social scene understanding. Peruzzi’s major contributions center on developing data-driven approaches to automatically detect and analyze conversational groups in images and videos, moving beyond traditional hand-crafted methods. His most cited work, “Improving Social Awareness Through DANTE: Deep Affinity Network for Clustering Conversational Interactants” (2020), introduces a novel deep learning architecture that predicts the likelihood of individuals belonging to the same conversational group by analyzing their spatial arrangements. This work, with 39 citations, has advanced the field’s ability to understand social dynamics in unconstrained environments. His earlier paper on “Conversational Group Detection With Deep Convolutional Networks” (2018) laid foundational work by building on the concept of o-spaces—social gathering spaces that help assign people to groups. Peruzzi’s research has significant implications for applications in video surveillance, social robotics, and human-computer interaction, making him a notable contributor to the growing field of computational social science.