PeopleofSuperintelligence

Researcher

Qixiang Ye

Also known as 叶齐祥, Ye Qixiang, Qi-Xiang Ye

Qixiang Ye is a distinguished professor at the University of Chinese Academy of Sciences (UCAS) and a coauthor of FreeAnchor, Conformer, VMamba and YOLOv12, methods for object detection and visual recognition. He earned bachelor's and master's degrees in mechanical and electrical engineering at Harbin Institute of Technology and a Ph.D. at the Institute of Computing Technology of the Chinese Academy of Sciences, was a visiting assistant professor at the University of Maryland, College Park until 2013, and was a visiting scholar at Duke University in 2016. He coauthored FreeAnchor at NeurIPS 2019, Conformer at ICCV 2021 and VMamba at NeurIPS 2024, and in 2025 he and coauthors Yunjie Tian and David Doermann released YOLOv12. In 2023, he won China's State Technological Innovation Award (2nd class) as a co-developer of a visual spatial intelligence system. He directs the Learning And Machine Perception (LAMP) lab at UCAS, where his research interests include visual spatial intelligence and self-evolution multi-modality models.

Timeline

  1. By 2016

    Degrees, Maryland and Duke

    He earns bachelor's and master's degrees in mechanical and electrical engineering at Harbin Institute of Technology and a Ph.D. at the Institute of Computing Technology, Chinese Academy of Sciences. He is a visiting assistant professor at UMIACS, the Institute for Advanced Computer Studies at the University of Maryland, College Park, until 2013, and a visiting scholar in electrical and computer engineering at Duke University in 2016.

  2. 2019

    FreeAnchor at NeurIPS

    He coauthors FreeAnchor with Xiaosong Zhang, Fang Wan, Chang Liu and Rongrong Ji, published at NeurIPS 2019. The method replaces hand-crafted anchor assignment in object detectors with learned anchor matching, formulated as maximum likelihood estimation.

  3. 2021

    Conformer, a hybrid backbone

    He coauthors Conformer with Zhiliang Peng, Wei Huang, Shanzhi Gu, Lingxi Xie, Yaowei Wang and Jianbin Jiao, presented at ICCV 2021. The hybrid network joins convolutional local features with transformer global representations through a Feature Coupling Unit and outperforms DeiT-B by 2.3% on ImageNet.

  4. 2023

    State Technological Innovation Award

    He wins the State Technological Innovation Award (2nd class) of China in 2023 as a co-developer of a visual spatial intelligence system.

  5. 2024

    VMamba, a vision backbone

    He coauthors VMamba with Yue Liu, Yunjie Tian, Yuzhong Zhao, Hongtian Yu, Lingxi Xie, Yaowei Wang, Jianbin Jiao and Yunfan Liu at NeurIPS 2024. The paper adapts Mamba, a state-space language model, into a vision backbone with linear time complexity, built from Visual State-Space blocks.

  6. 2025

    YOLOv12, an attention-centric detector

    In February 2025, he and coauthors Yunjie Tian and David Doermann post YOLOv12, an attention-centric YOLO framework that matches the speed of earlier CNN-based versions. YOLOv12-N reaches 40.6% mAP at 1.64 ms latency on a T4 GPU, and the work is a NeurIPS 2025 paper.

  7. Since 2022

    Distinguished professor at UCAS

    He is a distinguished professor at the University of Chinese Academy of Sciences since 2022 and directs its Learning And Machine Perception (LAMP) lab. His research interests include visual spatial intelligence and self-evolution multi-modality models. He is an area chair for CVPR 2023, ICLR 2024–25 and NeurIPS 2024–25, and was on the editorial boards of IEEE Transactions on Intelligent Transportation Systems and IEEE Transactions on Circuits and Systems for Video Technology.

Career

  1. Since 2022University of Chinese Academy of SciencesDistinguished Professor
  2. 2016Duke UniversityVisiting Scholar
  3. Until 2013University of MarylandVisiting Assistant Professor

Education

  1. Harbin Institute of TechnologyBachelor's degree
  2. Chinese Academy of SciencesPhD

From the biography.

  1. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes University of Chinese Academy of Sciences. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.
  2. The median h-index is 48 at universities and 14 at frontier AI labsIncludes University of Chinese Academy of Sciences. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.
  3. Google has the highest combined h-index among AI organizationsIncludes University of Chinese Academy of Sciences. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  4. Companies supply the most prior-work hires at 7 of the 9 frontier AI labsIncludes University of Chinese Academy of Sciences. Our estimates of the organizations where 550 people worked directly before joining 9 frontier AI labs across 620 career transitions, comparing the main feeder organizations per lab.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Qixiang Ye yet. Send us a free-licensed photo.

  1. 叶齐祥-中国科学院大学-UCAS people.ucas.ac.cn. Checked 22 October 2026.
  2. FreeAnchor: Learning to Match Anchors for Visual Object Detection papers.nips.cc. Checked 22 October 2026.
  3. Conformer: Local Features Coupling Global Representations for Visual Recognition (ICCV 2021 Open Access Repository) openaccess.thecvf.com. Checked 22 October 2026.
  4. VMamba: Visual State Space Model proceedings.neurips.cc. Checked 22 October 2026.
  5. YOLOv12: Attention-Centric Real-Time Object Detectors arxiv.org. Checked 22 October 2026.
  6. GitHub - sunsmarterjie/yolov12: [NeurIPS 2025] YOLOv12: Attention-Centric Real-Time Object Detectors github.com. Checked 22 October 2026.