Monica Haurilet

Papers

Total Citations

H-Index

About

Monica Haurilet is a researcher whose work lies at the intersection of computer vision and natural language processing, with a particular focus on visual question answering (VQA) and scene understanding. Her key contributions center on developing structured, interpretable representations of visual data to enable more sophisticated reasoning in AI systems. In her notable work, "DynGraph: Visual Question Answering via Dynamic Scene Graphs" (2019), Haurilet introduced a novel approach that leverages dynamic scene graphs—structured representations of objects, attributes, and relationships in an image—to answer complex visual questions. This method moves beyond static, feature-based models by capturing temporal and relational changes, allowing for more nuanced and context-aware reasoning. While her citation count is modest, with the DynGraph paper receiving 2 citations, her work represents a forward-thinking step in integrating graph-based reasoning with VQA, a field that continues to grow in importance. Haurilet’s research is particularly valuable for students and researchers interested in bridging the gap between low-level visual features and high-level semantic understanding, offering a framework that could inspire future work in interactive and dynamic visual AI systems.