Reza Bosagh Zadeh
Papers
1
Total Citations
168
H-Index
1
About
Reza Bosagh Zadeh is a leading researcher at the intersection of machine learning, 3D computer vision, and large-scale distributed systems. His most celebrated contribution is **FusionNet**, a pioneering deep learning architecture that achieved state-of-the-art results on the Princeton ModelNet challenge by fusing volumetric and point-based 3D data representations. This work, which has garnered **168 citations**, fundamentally advanced how machines perceive and classify 3D objects, directly impacting autonomous robotics and augmented reality. Beyond vision, Zadeh is renowned for his foundational work in scalable machine learning, particularly for developing the **AllReduce** algorithm used in the **Spark** distributed computing framework—a core technology enabling massive parallel training of models across thousands of machines. His research has been published in top-tier venues including NeurIPS, ICML, and CVPR, and his algorithms are deployed in production systems handling petabytes of data. A former Twitter engineer and current entrepreneur, Zadeh’s work bridges rigorous theoretical foundations with practical, high-impact systems, making him a pivotal figure in both the academic and industrial AI communities.
Research Focus
Key Achievements
Top Papers
- 1FusionNet: 3D Object Classification Using Multiple Data Representations168 citations · 2016