Xiaoyan Zhao
Papers
2
Total Citations
7
H-Index
2
About
Xiaoyan Zhao’s research bridges the critical gap between computer vision and acoustic signal processing, with a focus on real-time, embedded systems. Her most cited work introduces a **two-stage Hough transform algorithm** for lane detection, optimized for the TMS320DM6437 DSP platform. By leveraging YCbCr color space to robustly extract white and yellow lane markings, this contribution directly enhances the reliability of autonomous vehicle navigation and industrial robotics, demonstrating a practical, hardware-efficient approach to a core perception challenge. In parallel, Zhao has advanced the field of **sound source localization (SSL)** for noisy, reverberant environments. Her 2022 paper proposes a **Convolutional Residual Network (CRN)** applied to microphone array data, significantly improving localization accuracy over traditional methods. This work has direct applications in video conferencing, robotic hearing, and speech enhancement, where robust spatial awareness is essential. With a growing citation footprint and a clear trajectory from embedded vision to deep learning-based acoustics, Zhao’s work exemplifies the integration of classical algorithms with modern neural architectures. Her contributions are particularly valuable for researchers developing low-latency, real-world perception systems for autonomous platforms and human-machine interaction.
Research Focus
Key Achievements
Top Papers
- 1
- 2