Xiaoying Song

About

Xiaoying Song’s research centers on computational vision and neural network architectures, with a particular focus on general object recognition and location in complex, real-world environments. Her most notable contribution is the development of the Where-What Network (WWN), a biologically inspired model that integrates object identification with spatial localization, addressing a fundamental challenge in machine perception. In her seminal 2011 paper, “Where-What Network with CUDA: General Object Recognition and Location in Complex Backgrounds,” Song demonstrated how parallel computing (CUDA) could accelerate the network’s performance, enabling robust recognition even amidst visual clutter. This work, which has garnered 3 citations, laid the groundwork for scalable, brain-like vision systems that can simultaneously answer “where” and “what” an object is. Though her citation count is modest, her research represents an early and influential step toward unifying recognition and localization—a problem that remains central to modern computer vision and robotics. Song’s work is particularly valuable for students and researchers exploring biologically plausible AI, offering a clear example of how neural network design can bridge perception and action in dynamic scenes.

Research Focus

Key Achievements

Top Papers

1. 1
   Where-What Network with CUDA: General Object Recognition and Location in Complex Backgrounds

   3 citations · 2011

Key Collaborators

Contact & Links