Heydar Maboudi Afkham

Papers

Total Citations

H-Index

About

Heydar Maboudi Afkham’s research focuses on 3D object categorization and computer vision, with a particular emphasis on extracting essential local object characteristics to improve recognition systems. His most-cited work, "Extracting essential local object characteristics for 3D object categorization" (2013), challenges traditional approaches that rely on summarizing local features through simple occurrence counts. Instead, Afkham proposes a method to identify and leverage only the most discriminative features shared across object classes, significantly enhancing categorization accuracy. This contribution addresses a fundamental bottleneck in 3D vision: the tendency for similar local appearances to obscure class distinctions. While his citation count is modest, his work demonstrates a thoughtful departure from conventional bag-of-features models, offering a more nuanced understanding of how local geometry and appearance can be distilled for efficient recognition. Afkham’s research is particularly relevant for students and researchers working on 3D object recognition, feature selection, and shape analysis, as it underscores the importance of focusing on essential, class-specific characteristics rather than exhaustive feature aggregation. His approach continues to inspire discussions on minimal yet effective representations in computer vision.