Yifan Guo

About

Yifan Guo’s research bridges the gap between high-performance deep learning and real-world deployment, with a focus on computer vision and acoustic signal processing. In their most-cited work, “QATFP-YOLO,” Guo tackles a critical challenge: bringing state-of-the-art object detection—essential for self-driving cars, surveillance, and robotics—to non-GPU devices. By integrating quantization-aware training with filter pruning, they significantly reduce model size and computational cost without sacrificing accuracy, enabling efficient YOLO-based detection on resource-constrained hardware. This contribution has already garnered 4 citations, reflecting its practical relevance for edge computing and embedded systems. Earlier, Guo explored spatial audio perception in “A robust high resolution speaker DOA estimation under reverberant environment,” where they combined acoustic vector sensors with spatial sparsity representation to improve direction-of-arrival estimation for service robots in noisy, echoic settings. This work, with 2 citations, highlights their versatility in both vision and audition domains. Guo’s research demonstrates a consistent commitment to making advanced AI techniques accessible and robust for real-world applications, from autonomous navigation to human-robot interaction.

Research Focus

Key Achievements

Top Papers

1. 1
   QATFP-YOLO: Optimizing Object Detection on Non-GPU Devices with YOLO Using Quantization-Aware Training and Filter Pruning

   4 citations · 2024
2. 2
   A robust high resolution speaker DOA estimation under reverberant environment

   2 citations · 2014

Key Collaborators

Contact & Links