Siyuan Qiao

Google (United States), Huazhong University of Science and Technology

Papers

Total Citations

H-Index

About

Siyuan Qiao is a leading researcher at the intersection of computer vision and autonomous driving, with a primary focus on semantic and panoptic segmentation. His work addresses the critical challenge of enabling machines to understand visual scenes at the pixel level—a fundamental requirement for safe and reliable autonomous navigation. Qiao’s most significant contribution is his pioneering work on the **Waymo Open Dataset: Panoramic Video Panoptic Segmentation**, which has garnered over 55 citations. This work provides a large-scale, high-quality benchmark for panoptic segmentation in autonomous driving, pushing the field toward more comprehensive scene understanding by unifying semantic class and instance identification. He further advanced efficient segmentation with his work on **Superpixel Transformers**, introducing a novel architecture that balances accuracy and computational efficiency for real-time applications. Beyond perception, Qiao has also explored motion planning, as seen in his recent work on **FDSPC**, which addresses smooth trajectory generation for mobile robots. His research is highly impactful, directly influencing both academic benchmarks and practical systems in robotics and autonomous driving, making him a key figure in modern machine perception.