Ziyu Zhao
Papers
1
Total Citations
6
H-Index
1
About
Ziyu Zhao is an emerging researcher specializing in 3D perception and efficient deep learning for autonomous systems. Their most notable work, "PillarHist: A Quantization-aware Pillar Feature Encoder based on Height-aware Histogram" (2025), demonstrates a focused expertise at the intersection of 3D object detection, model quantization, and real-time inference for autonomous driving and robotics applications. This contribution addresses a critical challenge in onboard deployment: maintaining high detection performance while reducing computational overhead through hardware-friendly quantization strategies applied to pillar-based detection architectures. By introducing a height-aware histogram approach to pillar feature encoding, Zhao's research advances the practical viability of deploying sophisticated 3D perception models on resource-constrained embedded platforms — a bottleneck that has long hindered real-world autonomous vehicle systems. The work has already garnered 6 citations shortly after publication in 2025, signaling meaningful early traction within the autonomous driving and efficient AI communities. Zhao represents a new generation of researchers bridging the gap between state-of-the-art deep learning methodology and the stringent efficiency demands of safety-critical real-world deployment, making their work particularly relevant to practitioners and academics working on scalable autonomous perception systems.
Research Focus
Key Achievements
Top Papers
- 1