PeopleofSuperintelligence

Researcher

Coenraad Mouton

Coenraad Mouton is a deep learning researcher who studies neural-network generalization and adversarial robustness. He coauthored early work on convolutional-network invariance and classification margins, completed a PhD at North-West University in 2024, and joined the University of Kiel's Intelligent Imaging Lab in October 2024. His later work tests adversarial-training methods and studies nonrobust features and fracture-risk prediction in medical images. As of October 2026, he works in the Section for Biomedical Imaging at Schleswig-Holstein University Hospital's Kiel campus.

Now at
Christian-Albrechts-Universität zu Kiel and Department of Radiology and Neuroradiology

Timeline

  1. 2020

    CNN translation invariance

    He coauthors a SACAIR paper on stride and translation invariance in convolutional neural networks in 2020.

  2. 2022

    The Missing Margin

    With Marthinus W. Theunissen and Marelie H. Davel, he studies how sample corruption changes classification margins in artificial neural networks. Their paper appears in SACAIR in 2022.

  3. 2024

    A PhD on generalization

    He completes a PhD at North-West University in 2024 on margin-based prediction of deep-network generalization. His paper ‘Input margins can predict generalization too’ appears at AAAI in 2024.

  4. October 2024

    Joining Kiel's imaging lab

    He joins the Intelligent Imaging Lab at the University of Kiel in October 2024.

  5. 2025

    Comparing adversarial training

    In 2025, he coauthors a study comparing two margin-maximizing adversarial-training methods for deep neural networks. The paper evaluates a loss-based method and Dynamics-Aware Robust Training.

  6. 2026

    Nonrobust features in medical images

    In 2026, he studies nonrobust features in medical-imaging networks. His experiments show that models trained only on these features achieve above-chance accuracy on five MedMNIST classification tasks.

  7. 2026

    Fracture-risk prediction from images

    In 2026, he coauthors the Attention-to-Survival Fusion model, which combines clinical risk factors with pelvic-radiograph features to estimate time-dependent fracture risk.

  8. As of October 2026

    Postdoctoral research in Kiel

    As of October 2026, he is a postdoc at the Intelligent Imaging Lab at the University of Kiel, where he works on medical-imaging applications related to osteoporosis. His research role is in the Section for Biomedical Imaging within the Department of Radiology and Neuroradiology at Schleswig-Holstein University Hospital's Kiel campus.

Career

  1. Since 2024Christian-Albrechts-Universität zu KielPostdoc
  2. NowDepartment of Radiology and Neuroradiology

Education

  1. 2024North-West UniversityPhD

From the biography.

  1. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes Christian-Albrechts-Universität zu Kiel. 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.
  2. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Christian-Albrechts-Universität zu Kiel. 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.
  3. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Christian-Albrechts-Universität zu Kiel. 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.
  4. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.

Colleagues

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

Sources and credits

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

  1. Publications | engineering.nwu.ac.za engineering.nwu.ac.za. Checked 21 October 2026.
  2. The Missing Margin: How Sample Corruption Affects Distance to the Boundary in ANNs arxiv.org. Checked 21 October 2026.
  3. Coenraad Mouton coenraadmouton.com. Checked 21 October 2026.
  4. On margin-based generalization prediction in deep neural networks arxiv.org. Checked 21 October 2026.
  5. Does simple trump complex? Comparing strategies for adversarial robustness in DNNs arxiv.org. Checked 21 October 2026.
  6. Useful nonrobust features are ubiquitous in biomedical images arxiv.org. Checked 21 October 2026.
  7. Attention-to-Survival: Multimodal Fracture Risk Prediction Based on Pelvic Radiographs and Clinical Data from the Study of Osteoporotic Fractures proceedings.mlr.press. Checked 21 October 2026.
  8. Dr. Coenraad Mouton univis.uni-kiel.de. Checked 21 October 2026.