PeopleofSuperintelligence

Researcher

Risheng Liu

Also known as Liu Risheng, Ri-Sheng Liu, 刘日升

Risheng Liu is a professor at Dalian University of Technology who studies machine learning, computer vision and numerical optimization, with work on bilevel optimization and low-light image enhancement. After earning a mathematics degree in 2007 and a doctorate in computational mathematics in 2012 at Dalian, he joined its School of Software in 2012, worked as a Hong Kong Scholar at the Hong Kong Polytechnic University from 2016 to 2017, and became a full professor in 2019. In 2021, he first-authored a survey of bi-level optimization for learning and vision and the Bi-level Descent Aggregation framework, and in 2022 he coauthored the Self-Calibrated Illumination method for low-light images, an oral paper at CVPR. In 2024, he first-authored an ICML paper on a single-loop, Hessian-free algorithm for nonconvex bi-level optimization. He leads the Vision, Learning & Optimization Group and is deputy dean of the School of Software.

Now at
Dalian University of Technology, International School of Information Science & Engineering and Dalian University of Technology–Ritsumeikan University International School of Information and Software

Timeline

  1. 2007–2012

    Mathematics degrees at Dalian

    He earns a bachelor's degree in mathematics in 2007 and a doctorate in computational mathematics in 2012 at Dalian University of Technology, advised by Zhixun Su and Zhouchen Lin. From 2010 to 2012, he does joint PhD research at Carnegie Mellon University's Robotics Institute, hosted by Fernando De la Torre.

  2. 2012–2015

    Joining DUT's software school

    In 2012, he joins the School of Software at Dalian University of Technology. He becomes an associate professor there in December 2015.

  3. 2016–2017

    Hong Kong Scholar at PolyU

    From 2016 to 2017, he does postdoctoral research as a Hong Kong Scholar in the computing department of the Hong Kong Polytechnic University, working with Professor Lei Zhang.

  4. 2019

    Excellent Young Scholar and professor

    The 2019 National Natural Science Fund for Excellent Young Scholars list, released in August 2019, includes him as an associate professor at Dalian University of Technology. He becomes a full professor there in November 2019.

  5. 2021

    Bilevel optimization for learning

    In 2021, he is first author of Investigating Bi-Level Optimization for Learning and Vision from a Unified Perspective: A Survey and Beyond, which expresses problems such as hyper-parameter optimization, meta-learning and neural architecture search as bi-level optimization and sets out a unified framework for gradient-based methods. He is also first author of A General Descent Aggregation Framework for Gradient-based Bi-level Optimization, which introduces the Bi-level Descent Aggregation (BDA) algorithmic framework.

  6. 2022

    Self-Calibrated Illumination for low light

    In 2022, he coauthors Toward Fast, Flexible, and Robust Low-Light Image Enhancement with Long Ma, Tengyu Ma, Xin Fan and Zhongxuan Luo. The paper introduces the Self-Calibrated Illumination (SCI) learning framework for brightening real-world low-light images, and it is an oral paper at CVPR 2022.

  7. 2024

    Moreau envelope for bilevel problems

    In 2024, he is first author of Moreau Envelope for Nonconvex Bi-Level Optimization: A Single-Loop and Hessian-Free Solution Strategy at the 41st International Conference on Machine Learning, with Zhu Liu, Wei Yao, Shangzhi Zeng and Jin Zhang. The paper gives a single-loop gradient-based algorithm built on a Moreau envelope reformulation, with convergence analysis for general nonconvex bi-level problems and tests on hyper-parameter learning and neural architecture search.

  8. As of October 2026

    Leading the VLOG group

    He leads the Vision, Learning & Optimization Group (VLOG) at the International School of Information Science & Engineering of Dalian University of Technology, whose research spans machine learning, computer vision and numerical optimization. He is also deputy dean of the School of Software and of the DUT–Ritsumeikan University International School of Information and Software. In July 2026, three papers from his group are accepted by IEEE TPAMI, and one more follows in August 2026.

Career

  1. Since 2012Dalian University of TechnologyFull Professor, Associate Professor
  2. NowInternational School of Information Science & EngineeringGroup Lead
  3. NowDalian University of Technology–Ritsumeikan University International School of Information and SoftwareDeputy Dean
  4. 2016–2017Hong Kong Polytechnic UniversityHong Kong Scholar
  5. 2010–2012Carnegie Mellon UniversityJoint PhD Researcher

Education

  1. 2012Dalian University of TechnologyDoctorate in Computational Mathematics
  2. 2007Dalian University of TechnologyBachelor's Degree in Mathematics

From the biography.

  1. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Dalian University of Technology. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  2. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes Dalian University of Technology. We estimate where the people shaping AI studied and where their careers led. Massachusetts Institute of Technology leads with 467 former students, including 291 recorded doctorates.
  3. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Dalian University of Technology. 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.
  4. Google has the highest combined h-index among AI organizationsIncludes Carnegie Mellon University. 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.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

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

  1. 大连理工大学主页平台管理系统 刘日升 首页 刘日升 faculty.dlut.edu.cn. Checked 22 October 2026.
  2. Risheng Liu | ZERO Lab zero-lab-pku.github.io. Checked 22 October 2026.
  3. 2019-Good News in 2019 ccfcv.ccf.org.cn. Checked 22 October 2026.
  4. Investigating Bi-Level Optimization for Learning and Vision from a Unified Perspective: A Survey and Beyond arxiv.org. Checked 22 October 2026.
  5. A General Descent Aggregation Framework for Gradient-based Bi-level Optimization arxiv.org. Checked 22 October 2026.
  6. Toward Fast, Flexible, and Robust Low-Light Image Enhancement arxiv.org. Checked 22 October 2026.
  7. Moreau Envelope for Nonconvex Bi-Level Optimization: A Single-Loop and Hessian-Free Solution Strategy proceedings.mlr.press. Checked 22 October 2026.
  8. VLOG@DUT rsliu.tech. Checked 22 October 2026.