idw – Informationsdienst Wissenschaft
Nachrichten, Termine, Experten
Nachrichten, Termine, Experten
idw - Informationsdienst
Wissenschaft
The 4D Quantum Computer Vision research group at the Max Planck Institute (MPI) for Informatics in Saarbrücken, Germany, is investigating the potential of quantum computing for computer-based image processing, also known as visual computing. A new hybrid model, developed together with partners from the University of Udine and the University of Naples Federico II, combines a so-called “quantum neural network” with a classical neural network. In tests on established metrics the model outperforms previous methods while using significantly fewer network connections and requiring a shorter training time.
Researching technologies that cannot yet keep up with the state of the art today, but have the potential to transform entire fields tomorrow—that is the aim of basic research. One such technology is quantum computing. Quantum computers are still rare, difficult to access, and in many applications slower than classical computers. At the same time, quantum computers promise major advantages thanks to their special properties: the ability to represent many quantum states simultaneously (superposition), and to link qubits into a coherent shared state so they can no longer be regarded as independent of one another (entanglement). Investigating where these properties can be beneficial is a new and active field of research.
At the MPI for Informatics, the 4D Quantum Computer Vision group led by Dr. Vladislav Golyanik is exploring how quantum computing could advance computer-based image processing, particularly in visual computing and 3D computer vision. “On the one hand, we look for alternative ‘quantum solution paths’ for problems that are already known. Beyond that, of course, we also look for entirely new research questions that can benefit from quantum computing,” explains Vladislav Golyanik. The goal is ambitious: in the future, the team aims to reconstruct arbitrary real-world scenes photorealistically on a computer from just a few recorded images—and from novel, unrecorded viewpoints—even when objects in the scene move or deform. Such tasks are extremely computationally intensive, pushing even modern high-performance computers to their limits.
To address this challenge, the group is currently working on so-called “quantum neural networks” (QNNs). In such models, quantum circuits, usually simulated on classical hardware, are treated as trainable functions, similar to classical machine learning, and are adjusted using data-driven training and optimization methods.
In a recently published paper titled “Quantum Visual Fields with Neural Amplitude Encoding,” presented in December 2025 at the Conference on Neural Information Processing Systems (NeurIPS), the team—consisting of PhD student Shuteng Wang, Dr. Vladislav Golyanik, and Professor Christian Theobalt, Scientific Director of the Visual Computing and Artificial Intelligence department at the MPI for Informatics—describes a method for processing image data so that QNNs can handle it more efficiently. To this end, the researchers developed an initial hybrid model that combines a classical neural network with a QNN to encode input signals in 2D or 3D, as well as collections of such signals.
Building on this, a follow-up project titled “QNeRF: Neural Radiance Fields on a Simulated Gate-based Quantum Computer” was carried out together with partners from the University of Udine and the University of Naples Federico II. In this project, the researchers developed a hybrid model for “novel-viewpoint rendering,” that is, synthesizing new viewpoints of objects or scenes that were not contained in the original data. In experiments, the new model achieved comparable or better quality on the established PSNR metric than previous approaches, while using less than half the required network parameters and requiring fewer training steps. “For our tests, we simulated the quantum circuits on classical hardware. That meant the absolute training time was higher, but we were able to reduce the number of training steps, which could indicate shorter training times on ‘real’ quantum hardware in the future,” explains Golyanik. “QNeRF” is currently publicly available as a preprint.
Additionally, the work of the research group 4D Quantum Computer Vision is funded by the German Research Foundation (DFG) through the project “Applied Quantum Computing for Computer Vision.”
Professor Christian Theobalt says: “For us, quantum computing is a bet on the future, a high-risk, high-gain approach: we want to understand early on which fundamental principles of quantum computing could truly be useful, so that we are scientifically and technologically prepared when quantum resources become more widely available.”
In the long term, the ability to reconstruct complex scenes more efficiently could enable applications in areas such as robotics, manufacturing, medicine, or film and visual effects—anywhere computers need to represent reliable 3D information from camera footage and compute novel viewpoints.
Editor and press contact:
Philipp Zapf-Schramm
Max-Planck-Institut für Informatik
Tel: +49 681 9325 4509
E-Mail: pzs@mpi-inf.mpg.de
Dr. Vladislav Golyanik
Research Group Leader, MPI for Informatics
Email: golyanik@mpi-inf.mpg.de
Prof. Dr. Christian Theobalt
Scientific Director, MPI for Informatics
Wang, Shuteng; Theobalt, Christian; and Golyanik, Vladislav. “Quantum Visual Fields with Neural Amplitude Encoding.” Neural Information Processing Systems (NeurIPS), 2025.
Project page: https://4dqv.mpi-inf.mpg.de/QVF/
Lizzio Bosco, Daniele; Wang, Shuteng; Serra, Giuseppe; and Golyanik, Vladislav. “QNeRF: Neural Radiance Fields on a Simulated Gate-based Quantum Computer.” arXiv preprint arXiv:2601.05250, 2026. Project page: https://4dqv.mpi-inf.mpg.de/QNeRF/
https://4dqv.mpi-inf.mpg.de/ Website 4D Quantum Computer Vision
https://www.mpi-inf.mpg.de/home Website MPI for Informatics
https://gepris.dfg.de/gepris/projekt/534951134?language=en DFG project Applied Quantum Computer Vision
Dr. Vladislav Golyanik, head of the research group 4D Quantum Computer Vision
Source: Philipp Zapf-Schramm
Copyright: Max Planck Institute for Informatics
Dr. Vladislav Golyanik, head of the research group 4D Quantum Computer Vision
Source: Philipp Zapf-Schramm
Copyright: Max Planck Institute for Informatics
Criteria of this press release:
Journalists, Scientists and scholars, all interested persons
Information technology
transregional, national
Research results
English
You can combine search terms with and, or and/or not, e.g. Philo not logy.
You can use brackets to separate combinations from each other, e.g. (Philo not logy) or (Psycho and logy).
Coherent groups of words will be located as complete phrases if you put them into quotation marks, e.g. “Federal Republic of Germany”.
You can also use the advanced search without entering search terms. It will then follow the criteria you have selected (e.g. country or subject area).
If you have not selected any criteria in a given category, the entire category will be searched (e.g. all subject areas or all countries).
